Allow memory_region_register_iommu_notifier() to fail
v4 cover Allow memory_region_register_iommu_notifier() to fail
This series allows the memory_region_register_iommu_notifier() to fail. As of now, when a MAP notifier is attempted to be registered along with SMMUv3 or AMD IOMMU, we exit in the IOMMU MR notify_flag_changed() callback.
In case of VFIO assigned device hotplug, this could be handled more nicely directly within the VFIO code, simply rejecting the hotplug without exiting. This is what the series achieves by handling the memory_region_register_iommu_notifier() returned value and Error object.
To propagate errors collected during vfio_listener_region_add() we now store the error handle inside the VFIO container instead of a returned value.
The message now is:
(QEMU) device_add id=hot0 driver=vfio-pci host=0000:89:00.0 bus=pcie.1 {"error": {"class": "GenericError", "desc": "vfio 0000:89:00.0: failed to setup container for group 2: memory listener initialization failed:
Region smmuv3-iommu-memory-region-0-6: device 01.00.0 requires iommu MAP notifier which is not currently supported"}}
Best Regards
Eric
This series can be found at:
https://github.com/eauger/qemu/tree/v4.1.0_register_iommu_notifier_fail_v4
History:
v3 -> v4:
- added Peter's R-b on 2d patch
- 1st patch: restore hw_error, remove useless ret assignment, improve DMA host window error message, remove local mr variable
v2 -> v3:
- also pass an Error handle (suggested by Peter)
v1 -> v2:
- Intel IOMMU now handles the problem differently with machine init done notifier and machine hotplug allowed hook.
- use assert(!ret)
- message rewording in SMMUv3
Follow-up of "VFIO/SMMUv3: Fail on VFIO/HW nested paging detection"
https://patchew.org/QEMU/[email protected]/
Eric Auger (2):
vfio: Turn the container error into an Error handle
memory: allow memory_region_register_iommu_notifier() to fail
exec.c | 10 +++++-- hw/arm/smmuv3.c | 18 ++++++------ hw/i386/amd_iommu.c | 17 +++++++----- hw/i386/intel_iommu.c | 8 ++++-- hw/ppc/spapr_iommu.c | 8 ++++-- hw/vfio/common.c | 52 +++++++++++++++++++++++------------ hw/vfio/spapr.c | 4 ++- hw/virtio/vhost.c | 9 ++++-- include/exec/memory.h | 21 ++++++++++---- include/hw/vfio/vfio-common.h | 2 +- memory.c | 31 +++++++++++++-------- 11 files changed, 120 insertions(+), 60 deletions(-)
Signature · 1 line
2.20.1
PATCH v4 1/2 vfio: Turn the container error into an Error handle
The container error integer field is currently used to store the first error potentially encountered during any vfio_listener_region_add() call. However this fails to propagate detailed error messages up to the vfio_connect_container caller. Instead of using an integer, let's use an Error handle.
Messages are slightly reworded to accomodate the propagation.
---
v3 -> v4:
- remove ret useless assignments and restore hw_error()
- remove mr local variable
- trace [start, end] instead of [start, size] and improve the message for overalp with existing DMA host window
--- hw/vfio/common.c | 43 +++++++++++++++++++++++------------ hw/vfio/spapr.c | 4 +++- include/hw/vfio/vfio-common.h | 2 +- 3 files changed, 32 insertions(+), 17 deletions(-)
hw/vfio/common.c +28 -15
@@ -509,6 +509,7 @@ static void vfio_listener_region_add(MemoryListener *listener,int ret;VFIOHostDMAWindow *hostwin;bool hostwin_found; +Error *err =NULL ;if (vfio_listener_skipped_section (section)) { trace_vfio_listener_region_add_skip(
@@ -543,13 +544,20 @@ static void vfio_listener_region_add(MemoryListener *listener, hostwin->max_iova - hostwin->min_iova + 1, section->offset_within_address_space, int128_get64(section->size))) { - ret =-1 ; +error_setg (&err, +"region [0x%" PRIx64",0x%" PRIx64"] overlaps with existing" +"host DMA window [0x%" PRIx64",0x%" PRIx64"]" , + section->offset_within_address_space, + section->offset_within_address_space + +int128_get64 (section->size) -1 , + hostwin->min_iova, hostwin->max_iova); goto fail; } } ret =vfio_spapr_create_window (container, section, &pgsize);if (ret) { +error_setg_errno (&err, -ret,"Failed to create SPAPR window" ); goto fail; }
@@ -594,10 +602,8 @@ static void vfio_listener_region_add(MemoryListener *listener, }if (!hostwin_found) { -error_report ("vfio: IOMMU container %p can't map guest IOVA region" -" 0x%" HWADDR_PRIx"..0x%" HWADDR_PRIx, - container, iova, end); - ret = -EFAULT ; +error_setg (&err,"Container %p can't map guest IOVA region" +" 0x%" HWADDR_PRIx"..0x%" HWADDR_PRIx, container, iova, end); goto fail; }
@@ -664,11 +670,12 @@ static void vfio_listener_region_add(MemoryListener *listener, ret =vfio_dma_map (container, iova,int128_get64 (llsize), vaddr, section->readonly);if (ret) { -error_report ("vfio_dma_map(%p, 0x%" HWADDR_PRIx", " -"0x%" HWADDR_PRIx", %p) = %d (%m)" , - container, iova,int128_get64 (llsize), vaddr, ret); +error_setg (&err,"vfio_dma_map(%p, 0x%" HWADDR_PRIx", " +"0x%" HWADDR_PRIx", %p) = %d (%m)" , + container, iova,int128_get64 (llsize), vaddr, ret);if (memory_region_is_ram_device (section->mr)) {/* Allow unexpected mappings not to be fatal for RAM devices */ +error_report_err (err);return ; } goto fail;
@@ -688,9 +695,14 @@ fail:*/ if (!container->initialized) {if (!container->error) { - container->error = ret; +error_propagate_prepend (&container->error, err, +"Region %s: " , +memory_region_name (section->mr)); + }else { +error_free (err); } }else { +error_report_err (err);hw_error ("vfio: DMA mapping failed, unable to continue" ); } }
@@ -1251,6 +1263,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container =g_malloc0 (sizeof (*container)); container->space = space; container->fd = fd; + container->error =NULL ;QLIST_INIT (&container->giommu_list);QLIST_INIT (&container->hostwin_list);
@@ -1308,9 +1321,9 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, &address_space_memory);if (container->error) {memory_listener_unregister (&container->prereg_listener); - ret = container->error; -error_setg (errp, -"RAM memory listener initialization failed for container" ); + ret =-1 ; +error_propagate_prepend (errp, container->error, +"RAM memory listener initialization failed: " ); goto free_container_exit; } }
@@ -1365,9 +1378,9 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,memory_listener_register (&container->listener, container->space->as);if (container->error) { - ret = container->error; -error_setg_errno (errp, -ret, -"memory listener initialization failed for container" ); + ret =-1 ; +error_propagate_prepend (errp, container->error, +"memory listener initialization failed: " ); goto listener_release_exit; }
hw/vfio/spapr.c +3 -1
@@ -17,6 +17,7 @@#include "hw/hw.h" #include "exec/ram_addr.h" #include "qemu/error-report.h" +#include "qapi/error.h" #include "trace.h" static bool vfio_prereg_listener_skipped_section (MemoryRegionSection *section)
@@ -85,7 +86,8 @@ static void vfio_prereg_listener_region_add(MemoryListener *listener,*/ if (!container->initialized) {if (!container->error) { - container->error = ret; +error_setg_errno (&container->error, -ret, +"Memory registering failed" ); } }else {hw_error ("vfio: Memory registering failed, unable to continue" );
include/hw/vfio/vfio-common.h +1 -1
@@ -71,7 +71,7 @@ typedef struct VFIOContainer {MemoryListener listener;MemoryListener prereg_listener;unsigned iommu_type; -int error; +Error *error;bool initialized;unsigned long pgsizes; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list;
Signature · 1 line
2.20.1
Re: [PATCH v4 1/2] vfio: Turn the container error into an Error handle
Quoted 164 lines · Eric Auger
The container error integer field is currently used to store the first error potentially encountered during any
vfio_listener_region_add()call. However this fails to propagate detailed error messages up to thevfio_connect_containercaller. Instead of using an integer, let's use an Error handle.Messages are slightly reworded to accomodate the propagation.
Signed-off-by Eric Auger <[email protected]>---
v3 -> v4:
- remove ret useless assignments and restorehw_error()
- remove mr local variable
- trace [start, end] instead of [start, size] and improve the message for overalp with existing DMA host window--- hw/vfio/common.c | 43 +++++++++++++++++++++++------------ hw/vfio/spapr.c | 4 +++- include/hw/vfio/vfio-common.h | 2 +- 3 files changed, 32 insertions(+), 17 deletions(-)hw/vfio/common.c +28
-15@@ -509,6 +509,7 @@ static void vfio_listener_region_add(MemoryListener *listener,int ret;VFIOHostDMAWindow *hostwin;bool hostwin_found; +Error *err =NULL ;if (vfio_listener_skipped_section (section)) { trace_vfio_listener_region_add_skip(@@ -543,13 +544,20 @@ static void vfio_listener_region_add(MemoryListener *listener, hostwin->max_iova - hostwin->min_iova + 1, section->offset_within_address_space, int128_get64(section->size))) { - ret =-1 ; +error_setg (&err, +"region [0x%" PRIx64",0x%" PRIx64"] overlaps with existing" +"host DMA window [0x%" PRIx64",0x%" PRIx64"]" , + section->offset_within_address_space, + section->offset_within_address_space + +int128_get64 (section->size) -1 , + hostwin->min_iova, hostwin->max_iova); goto fail; } } ret =vfio_spapr_create_window (container, section, &pgsize);if (ret) { +error_setg_errno (&err, -ret,"Failed to create SPAPR window" ); goto fail; }@@ -594,10 +602,8 @@ static void vfio_listener_region_add(MemoryListener *listener, }if (!hostwin_found) { -error_report ("vfio: IOMMU container %p can't map guest IOVA region" -" 0x%" HWADDR_PRIx"..0x%" HWADDR_PRIx, - container, iova, end); - ret = -EFAULT ; +error_setg (&err,"Container %p can't map guest IOVA region" +" 0x%" HWADDR_PRIx"..0x%" HWADDR_PRIx, container, iova, end); goto fail; }@@ -664,11 +670,12 @@ static void vfio_listener_region_add(MemoryListener *listener, ret =vfio_dma_map (container, iova,int128_get64 (llsize), vaddr, section->readonly);if (ret) { -error_report ("vfio_dma_map(%p, 0x%" HWADDR_PRIx", " -"0x%" HWADDR_PRIx", %p) = %d (%m)" , - container, iova,int128_get64 (llsize), vaddr, ret); +error_setg (&err,"vfio_dma_map(%p, 0x%" HWADDR_PRIx", " +"0x%" HWADDR_PRIx", %p) = %d (%m)" , + container, iova,int128_get64 (llsize), vaddr, ret);if (memory_region_is_ram_device (section->mr)) {/* Allow unexpected mappings not to be fatal for RAM devices */ +error_report_err (err);return ; } goto fail;@@ -688,9 +695,14 @@ fail:*/ if (!container->initialized) {if (!container->error) { - container->error = ret; +error_propagate_prepend (&container->error, err, +"Region %s: " , +memory_region_name (section->mr)); + }else { +error_free (err); } }else { +error_report_err (err);hw_error ("vfio: DMA mapping failed, unable to continue" ); } }@@ -1251,6 +1263,7 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, container =g_malloc0 (sizeof (*container)); container->space = space; container->fd = fd; + container->error =NULL ;QLIST_INIT (&container->giommu_list);QLIST_INIT (&container->hostwin_list);@@ -1308,9 +1321,9 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as, &address_space_memory);if (container->error) {memory_listener_unregister (&container->prereg_listener); - ret = container->error; -error_setg (errp, -"RAM memory listener initialization failed for container" ); + ret =-1 ; +error_propagate_prepend (errp, container->error, +"RAM memory listener initialization failed: " ); goto free_container_exit; } }@@ -1365,9 +1378,9 @@ static int vfio_connect_container(VFIOGroup *group, AddressSpace *as,memory_listener_register (&container->listener, container->space->as);if (container->error) { - ret = container->error; -error_setg_errno (errp, -ret, -"memory listener initialization failed for container" ); + ret =-1 ; +error_propagate_prepend (errp, container->error, +"memory listener initialization failed: " ); goto listener_release_exit; }hw/vfio/spapr.c +3
-1@@ -17,6 +17,7 @@#include "hw/hw.h" #include "exec/ram_addr.h" #include "qemu/error-report.h" +#include "qapi/error.h" #include "trace.h" static bool vfio_prereg_listener_skipped_section (MemoryRegionSection *section)@@ -85,7 +86,8 @@ static void vfio_prereg_listener_region_add(MemoryListener *listener,*/ if (!container->initialized) {if (!container->error) { - container->error = ret; +error_setg_errno (&container->error, -ret, +"Memory registering failed" ); } }else {hw_error ("vfio: Memory registering failed, unable to continue" );include/hw/vfio/vfio-common.h +0
-0
include/hw/vfio/vfio-common.h typedef struct VFIOContainer { · line 77
Quoted 8 lines
include/hw/vfio/vfio-common.h +1
-1@@ -71,7 +71,7 @@ typedef struct VFIOContainer {MemoryListener listener;MemoryListener prereg_listener;unsigned iommu_type; -int error; +Error *error;bool initialized;unsigned long pgsizes; QLIST_HEAD(, VFIOGuestIOMMU) giommu_list;
PATCH v4 2/2 memory: allow memory_region_register_iommu_notifier() to fail
In reply to Eric AugerCurrently, when a notifier is attempted to be registered and its flags are not supported (especially the MAP one) by the IOMMU MR, we generally abruptly exit in the IOMMU code. The failure could be handled more nicely in the caller and especially in the VFIO code.
So let's allow memory_region_register_iommu_notifier() to fail as well as notify_flag_changed() callback.
All sites implementing the callback are updated. This patch does not yet remove the exit(1) in the amd_iommu code.
in SMMUv3 we turn the warning message into an error message saying that the assigned device would not work properly.
---
v2 -> v3:
- turn the warning message into an error message
--- exec.c | 10 ++++++++-- hw/arm/smmuv3.c | 18 ++++++++++-------- hw/i386/amd_iommu.c | 17 ++++++++++------- hw/i386/intel_iommu.c | 8 +++++--- hw/ppc/spapr_iommu.c | 8 +++++--- hw/vfio/common.c | 9 +++++++-- hw/virtio/vhost.c | 9 +++++++-- include/exec/memory.h | 21 ++++++++++++++++----- memory.c | 31 ++++++++++++++++++++----------- 9 files changed, 88 insertions(+), 43 deletions(-)
exec.c +8 -2
@@ -663,7 +663,8 @@ static void tcg_register_iommu_notifier(CPUState *cpu,*/ MemoryRegion *mr =MEMORY_REGION (iommu_mr);TCGIOMMUNotifier *notifier; -int i; +Error *err =NULL ; +int i, ret;for (i =0 ; i < cpu->iommu_notifiers->len; i++) { notifier = g_array_index(cpu->iommu_notifiers, TCGIOMMUNotifier *, i);
@@ -692,7 +693,12 @@ static void tcg_register_iommu_notifier(CPUState *cpu,0 ,HWADDR_MAX , iommu_idx); -memory_region_register_iommu_notifier (notifier->mr, ¬ifier->n); + ret =memory_region_register_iommu_notifier (notifier->mr, ¬ifier->n, + &err); +if (ret) { +error_report_err (err); +exit (1 ); + } }if (!notifier->active) {
hw/arm/smmuv3.c +10 -8
@@ -1469,20 +1469,21 @@ static void smmuv3_class_init(ObjectClass *klass, void *data) dc->realize = smmu_realize; } -static void smmuv3_notify_flag_changed (IOMMUMemoryRegion *iommu, -IOMMUNotifierFlag old, -IOMMUNotifierFlag new) +static int smmuv3_notify_flag_changed (IOMMUMemoryRegion *iommu, +IOMMUNotifierFlag old, +IOMMUNotifierFlag new, +Error **errp) {SMMUDevice *sdev =container_of (iommu, SMMUDevice, iommu);SMMUv3State *s3 = sdev->smmu;SMMUState *s = &(s3->smmu_state);if (new &IOMMU_NOTIFIER_MAP ) { -int bus_num =pci_bus_num (sdev->bus); -PCIDevice *pcidev =pci_find_device (sdev->bus, bus_num, sdev->devfn); - -warn_report ("SMMUv3 does not support notification on MAP: " -"device %s will not function properly" , pcidev->name); +error_setg (errp, +"device %02x.%02x.%x requires iommu MAP notifier which is " +"not currently supported" ,pci_bus_num (sdev->bus), +PCI_SLOT (sdev->devfn),PCI_FUNC (sdev->devfn)); +return -EINVAL ; }if (old ==IOMMU_NOTIFIER_NONE ) {
@@ -1492,6 +1493,7 @@ static void smmuv3_notify_flag_changed(IOMMUMemoryRegion *iommu,trace_smmuv3_notify_flag_del (iommu->parent_obj.name);QLIST_REMOVE (sdev, next); } +return 0 ; } static void smmuv3_iommu_memory_region_class_init(ObjectClass *klass,
hw/i386/amd_iommu.c +10 -7
@@ -1466,18 +1466,21 @@ static const MemoryRegionOps mmio_mem_ops = { } }; -static void amdvi_iommu_notify_flag_changed (IOMMUMemoryRegion *iommu, -IOMMUNotifierFlag old, -IOMMUNotifierFlag new) +static int amdvi_iommu_notify_flag_changed (IOMMUMemoryRegion *iommu, +IOMMUNotifierFlag old, +IOMMUNotifierFlag new, +Error **errp) {AMDVIAddressSpace *as =container_of (iommu, AMDVIAddressSpace, iommu);if (new &IOMMU_NOTIFIER_MAP ) { -error_report ("device %02x.%02x.%x requires iommu notifier which is not " -"currently supported" , as->bus_num,PCI_SLOT (as->devfn), -PCI_FUNC (as->devfn)); -exit (1 ); +error_setg (errp, +"device %02x.%02x.%x requires iommu notifier which is not " +"currently supported" , as->bus_num,PCI_SLOT (as->devfn), +PCI_FUNC (as->devfn)); +return -EINVAL ; } +return 0 ; }static void amdvi_init (AMDVIState *s)
hw/i386/intel_iommu.c +5 -3
@@ -2929,9 +2929,10 @@ static IOMMUTLBEntry vtd_iommu_translate(IOMMUMemoryRegion *iommu, hwaddr addr,return iotlb; } -static void vtd_iommu_notify_flag_changed (IOMMUMemoryRegion *iommu, -IOMMUNotifierFlag old, -IOMMUNotifierFlag new) +static int vtd_iommu_notify_flag_changed (IOMMUMemoryRegion *iommu, +IOMMUNotifierFlag old, +IOMMUNotifierFlag new, +Error **errp) {VTDAddressSpace *vtd_as =container_of (iommu, VTDAddressSpace, iommu);IntelIOMMUState *s = vtd_as->iommu_state;
@@ -2944,6 +2945,7 @@ static void vtd_iommu_notify_flag_changed(IOMMUMemoryRegion *iommu, }else if (new ==IOMMU_NOTIFIER_NONE ) {QLIST_REMOVE (vtd_as, next); } +return 0 ; }static int vtd_post_load (void *opaque,int version_id)
hw/ppc/spapr_iommu.c +5 -3
@@ -205,9 +205,10 @@ static int spapr_tce_get_attr(IOMMUMemoryRegion *iommu,return -EINVAL ; } -static void spapr_tce_notify_flag_changed (IOMMUMemoryRegion *iommu, -IOMMUNotifierFlag old, -IOMMUNotifierFlag new) +static int spapr_tce_notify_flag_changed (IOMMUMemoryRegion *iommu, +IOMMUNotifierFlag old, +IOMMUNotifierFlag new, +Error **errp) {struct SpaprTceTable *tbl =container_of (iommu, SpaprTceTable, iommu);
@@ -216,6 +217,7 @@ static void spapr_tce_notify_flag_changed(IOMMUMemoryRegion *iommu, }else if (old !=IOMMU_NOTIFIER_NONE && new ==IOMMU_NOTIFIER_NONE ) {spapr_tce_set_need_vfio (tbl, false); } +return 0 ; }static int spapr_tce_table_post_load (void *opaque,int version_id)
hw/vfio/common.c +7 -2
@@ -636,9 +636,14 @@ static void vfio_listener_region_add(MemoryListener *listener, section->offset_within_region,int128_get64 (llend), iommu_idx); -QLIST_INSERT_HEAD (&container->giommu_list, giommu, giommu_next); -memory_region_register_iommu_notifier (section->mr, &giommu->n); + ret =memory_region_register_iommu_notifier (section->mr, &giommu->n, + &err); +if (ret) { +g_free (giommu); + goto fail; + } +QLIST_INSERT_HEAD (&container->giommu_list, giommu, giommu_next);memory_region_iommu_replay (giommu->iommu, &giommu->n);return ;
hw/virtio/vhost.c +7 -2
@@ -672,8 +672,9 @@ static void vhost_iommu_region_add(MemoryListener *listener, iommu_listener);struct vhost_iommu *iommu;Int128 end; -int iommu_idx; +int iommu_idx, ret;IOMMUMemoryRegion *iommu_mr; +Error *err =NULL ;if (!memory_region_is_iommu (section->mr)) {return ;
@@ -696,7 +697,11 @@ static void vhost_iommu_region_add(MemoryListener *listener, iommu->iommu_offset = section->offset_within_address_space - section->offset_within_region; iommu->hdev = dev; -memory_region_register_iommu_notifier (section->mr, &iommu->n); + ret =memory_region_register_iommu_notifier (section->mr, &iommu->n, &err); +if (ret) { +error_report_err (err); +exit (1 ); + }QLIST_INSERT_HEAD (&dev->iommu_list, iommu, iommu_next);/* TODO: can replay help performance here? */ }
include/exec/memory.h +16 -5
@@ -288,10 +288,16 @@ typedef struct IOMMUMemoryRegionClass {* @iommu: the IOMMUMemoryRegion * @old_flags: events which previously needed to be notified * @new_flags: events which now need to be notified +* +* Returns 0 on success, or a negative errno; in particular +* returns -EINVAL if the new flag bitmap is not supported by the +* IOMMU memory region. In case of failure, the error object +* must be created */ -void (*notify_flag_changed)(IOMMUMemoryRegion *iommu, -IOMMUNotifierFlag old_flags, -IOMMUNotifierFlag new_flags); +int (*notify_flag_changed)(IOMMUMemoryRegion *iommu, +IOMMUNotifierFlag old_flags, +IOMMUNotifierFlag new_flags, +Error **errp);/* Called to handle memory_region_iommu_replay(). * * The default implementation of memory_region_iommu_replay() is to
@@ -1067,13 +1073,18 @@ void memory_region_notify_one(IOMMUNotifier *notifier,* memory_region_register_iommu_notifier: register a notifier for changes to * IOMMU translation entries. * +* Returns 0 on success, or a negative errno otherwise. In particular, +* -EINVAL indicates that at least one of the attributes of the notifier +* is not supported (flag/range) by the IOMMU memory region. In case of error +* the error object must be created. +* * @mr: the memory region to observe * @n: the IOMMUNotifier to be added; the notify callback receives a * pointer to an #IOMMUTLBEntry as the opaque value; the pointer * ceases to be valid on exit from the notifier. */ -void memory_region_register_iommu_notifier (MemoryRegion *mr, -IOMMUNotifier *n); +int memory_region_register_iommu_notifier (MemoryRegion *mr, +IOMMUNotifier *n,Error **errp);/** * memory_region_iommu_replay: replay existing IOMMU translations to
memory.c +20 -11
@@ -1837,33 +1837,38 @@ bool memory_region_is_logging(MemoryRegion *mr, uint8_t client)return memory_region_get_dirty_log_mask (mr) & (1 << client); } -static void memory_region_update_iommu_notify_flags (IOMMUMemoryRegion *iommu_mr) +static int memory_region_update_iommu_notify_flags (IOMMUMemoryRegion *iommu_mr, +Error **errp) {IOMMUNotifierFlag flags =IOMMU_NOTIFIER_NONE ;IOMMUNotifier *iommu_notifier;IOMMUMemoryRegionClass *imrc =IOMMU_MEMORY_REGION_GET_CLASS (iommu_mr); +int ret =0 ; IOMMU_NOTIFIER_FOREACH(iommu_notifier, iommu_mr) { flags |= iommu_notifier->notifier_flags; }if (flags != iommu_mr->iommu_notify_flags && imrc->notify_flag_changed) { - imrc->notify_flag_changed (iommu_mr, - iommu_mr->iommu_notify_flags, - flags); + ret = imrc->notify_flag_changed (iommu_mr, + iommu_mr->iommu_notify_flags, + flags, errp); } - iommu_mr->iommu_notify_flags = flags; +if (!ret) { + iommu_mr->iommu_notify_flags = flags; + } +return ret; } -void memory_region_register_iommu_notifier (MemoryRegion *mr, -IOMMUNotifier *n) +int memory_region_register_iommu_notifier (MemoryRegion *mr, +IOMMUNotifier *n,Error **errp) {IOMMUMemoryRegion *iommu_mr; +int ret;if (mr->alias) { -memory_region_register_iommu_notifier (mr->alias, n); -return ; +return memory_region_register_iommu_notifier (mr->alias, n, errp); }/* We need to register for at least one bitfield */
@@ -1874,7 +1879,11 @@ void memory_region_register_iommu_notifier(MemoryRegion *mr, n->iommu_idx < memory_region_iommu_num_indexes(iommu_mr));QLIST_INSERT_HEAD (&iommu_mr->iommu_notify, n, node); -memory_region_update_iommu_notify_flags (iommu_mr); + ret =memory_region_update_iommu_notify_flags (iommu_mr, errp); +if (ret) { +QLIST_REMOVE (n, node); + } +return ret; }uint64_t memory_region_iommu_get_min_page_size (IOMMUMemoryRegion *iommu_mr)
@@ -1927,7 +1936,7 @@ void memory_region_unregister_iommu_notifier(MemoryRegion *mr, }QLIST_REMOVE (n, node); iommu_mr =IOMMU_MEMORY_REGION (mr); -memory_region_update_iommu_notify_flags (iommu_mr); +memory_region_update_iommu_notify_flags (iommu_mr,NULL ); } void memory_region_notify_one(IOMMUNotifier *notifier,
Signature · 1 line
2.20.1
Quoted 15 lines · Eric Auger
This series allows the
memory_region_register_iommu_notifier()to fail. As of now, when a MAP notifier is attempted to be registered along with SMMUv3 or AMD IOMMU, we exit in the IOMMU MRnotify_flag_changed()callback.In case of VFIO assigned device hotplug, this could be handled more nicely directly within the VFIO code, simply rejecting the hotplug without exiting. This is what the series achieves by handling the
memory_region_register_iommu_notifier()returned value and Error object.To propagate errors collected during
vfio_listener_region_add()we now store the error handle inside the VFIO container instead of a returned value.
Quoted 8 lines
The message now is:
(QEMU)device_addid=hot0 driver=vfio-pci host=0000:89:00.0 bus=pcie.1 {"error": {"class": "GenericError", "desc": "vfio 0000:89:00.0: failed to setup container for group 2: memory listener initialization failed:
Region smmuv3-iommu-memory-region-0-6: device 01.00.0 requires iommu MAP notifier which is not currently supported"}}Best Regards
For the series
and ppc parts
Quoted 43 lines
Eric
This series can be found at:
https://github.com/eauger/qemu/tree/v4.1.0_register_iommu_notifier_fail_v4History:
v3 -> v4:
- added Peter's R-b on 2d patch
- 1st patch: restorehw_error, remove useless ret assignment, improve DMA host window error message, remove local mr variablev2 -> v3:
- also pass an Error handle (suggested by Peter)v1 -> v2:
- Intel IOMMU now handles the problem differently with machine init done notifier and machine hotplug allowed hook.
- use assert(!ret)
- message rewording in SMMUv3Follow-up of "VFIO/SMMUv3: Fail on VFIO/HW nested paging detection"
https://patchew.org/QEMU/[email protected]/Eric Auger (2):
vfio: Turn the container error into an Error handle
memory: allowmemory_region_register_iommu_notifier()to failexec.c | 10 +++++-- hw/arm/smmuv3.c | 18 ++++++------ hw/i386/amd_iommu.c | 17 +++++++----- hw/i386/intel_iommu.c | 8 ++++-- hw/ppc/spapr_iommu.c | 8 ++++-- hw/vfio/common.c | 52 +++++++++++++++++++++++------------ hw/vfio/spapr.c | 4 ++- hw/virtio/vhost.c | 9 ++++-- include/exec/memory.h | 21 ++++++++++---- include/hw/vfio/vfio-common.h | 2 +- memory.c | 31 +++++++++++++-------- 11 files changed, 120 insertions(+), 60 deletions(-)
Signature · 4 lines
David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson
Quoted 53 lines · On Tue, 24 Sep 2019 10:25:15 +0200 Eric Auger <[email protected]> wrote:
This series allows the
memory_region_register_iommu_notifier()to fail. As of now, when a MAP notifier is attempted to be registered along with SMMUv3 or AMD IOMMU, we exit in the IOMMU MRnotify_flag_changed()callback.In case of VFIO assigned device hotplug, this could be handled more nicely directly within the VFIO code, simply rejecting the hotplug without exiting. This is what the series achieves by handling the
memory_region_register_iommu_notifier()returned value and Error object.To propagate errors collected during
vfio_listener_region_add()we now store the error handle inside the VFIO container instead of a returned value.The message now is:
(QEMU)device_addid=hot0 driver=vfio-pci host=0000:89:00.0 bus=pcie.1 {"error": {"class": "GenericError", "desc": "vfio 0000:89:00.0: failed to setup container for group 2: memory listener initialization failed:
Region smmuv3-iommu-memory-region-0-6: device 01.00.0 requires iommu MAP notifier which is not currently supported"}}Best Regards
Eric
This series can be found at:
https://github.com/eauger/qemu/tree/v4.1.0_register_iommu_notifier_fail_v4History:
v3 -> v4:
- added Peter's R-b on 2d patch
- 1st patch: restorehw_error, remove useless ret assignment, improve DMA host window error message, remove local mr variablev2 -> v3:
- also pass an Error handle (suggested by Peter)v1 -> v2:
- Intel IOMMU now handles the problem differently with machine init done notifier and machine hotplug allowed hook.
- use assert(!ret)
- message rewording in SMMUv3Follow-up of "VFIO/SMMUv3: Fail on VFIO/HW nested paging detection"
https://patchew.org/QEMU/[email protected]/Eric Auger (2):
vfio: Turn the container error into an Error handle
memory: allowmemory_region_register_iommu_notifier()to fail
Quoted 12 lines
exec.c | 10 +++++-- hw/arm/smmuv3.c | 18 ++++++------ hw/i386/amd_iommu.c | 17 +++++++----- hw/i386/intel_iommu.c | 8 ++++-- hw/ppc/spapr_iommu.c | 8 ++++-- hw/vfio/common.c | 52 +++++++++++++++++++++++------------ hw/vfio/spapr.c | 4 ++- hw/virtio/vhost.c | 9 ++++-- include/exec/memory.h | 21 ++++++++++---- include/hw/vfio/vfio-common.h | 2 +- memory.c | 31 +++++++++++++-------- 11 files changed, 120 insertions(+), 60 deletions(-)
For series,
Paolo, would this go in through you given the memory API changes and greater girth in patch 2/2? Thanks,
Alex
Quoted 69 lines · Alex Williamson
Quoted 53 lines · On Tue, 24 Sep 2019 10:25:15 +0200 Eric Auger <[email protected]> wrote:
This series allows the
memory_region_register_iommu_notifier()to fail. As of now, when a MAP notifier is attempted to be registered along with SMMUv3 or AMD IOMMU, we exit in the IOMMU MRnotify_flag_changed()callback.In case of VFIO assigned device hotplug, this could be handled more nicely directly within the VFIO code, simply rejecting the hotplug without exiting. This is what the series achieves by handling the
memory_region_register_iommu_notifier()returned value and Error object.To propagate errors collected during
vfio_listener_region_add()we now store the error handle inside the VFIO container instead of a returned value.The message now is:
(QEMU)device_addid=hot0 driver=vfio-pci host=0000:89:00.0 bus=pcie.1 {"error": {"class": "GenericError", "desc": "vfio 0000:89:00.0: failed to setup container for group 2: memory listener initialization failed:
Region smmuv3-iommu-memory-region-0-6: device 01.00.0 requires iommu MAP notifier which is not currently supported"}}Best Regards
Eric
This series can be found at:
https://github.com/eauger/qemu/tree/v4.1.0_register_iommu_notifier_fail_v4History:
v3 -> v4:
- added Peter's R-b on 2d patch
- 1st patch: restorehw_error, remove useless ret assignment, improve DMA host window error message, remove local mr variablev2 -> v3:
- also pass an Error handle (suggested by Peter)v1 -> v2:
- Intel IOMMU now handles the problem differently with machine init done notifier and machine hotplug allowed hook.
- use assert(!ret)
- message rewording in SMMUv3Follow-up of "VFIO/SMMUv3: Fail on VFIO/HW nested paging detection"
https://patchew.org/QEMU/[email protected]/Eric Auger (2):
vfio: Turn the container error into an Error handle
memory: allowmemory_region_register_iommu_notifier()to failQuoted 12 lines
exec.c | 10 +++++-- hw/arm/smmuv3.c | 18 ++++++------ hw/i386/amd_iommu.c | 17 +++++++----- hw/i386/intel_iommu.c | 8 ++++-- hw/ppc/spapr_iommu.c | 8 ++++-- hw/vfio/common.c | 52 +++++++++++++++++++++++------------ hw/vfio/spapr.c | 4 ++- hw/virtio/vhost.c | 9 ++++-- include/exec/memory.h | 21 ++++++++++---- include/hw/vfio/vfio-common.h | 2 +- memory.c | 31 +++++++++++++-------- 11 files changed, 120 insertions(+), 60 deletions(-)
Quoted 6 lines
For series,
Acked-by Alex Williamson <[email protected]>Paolo, would this go in through you given the memory API changes and greater girth in patch 2/2? Thanks,
Fine, queued.
Paolo