* [PULL 01/27] hw/cxl: Support aborting background commands
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 02/27] hw/cxl: Support get/set mctp response payload size Michael S. Tsirkin
` (26 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Davidlohr Bueso, Ajay Joshi, Jonathan Cameron,
Fan Ni
From: Davidlohr Bueso <dave@stgolabs.net>
As of 3.1 spec, background commands can be canceled with a new
abort command. Implement the support, which is advertised in
the CEL. No ad-hoc context undoing is necessary as all the
command logic of the running bg command is done upon completion.
Arbitrarily, the on-going background cmd will not be aborted if
already at least 85% done;
A mutex is introduced to stabilize mbox request cancel command vs
the timer callback being fired scenarios (as well as reading the
mbox registers). While some operations under critical regions
may be unnecessary (irq notifying, cmd callbacks), this is not
a path where performance is important, so simplicity is preferred.
Tested-by: Ajay Joshi <ajay.opensrc@micron.com>
Reviewed-by: Ajay Joshi <ajay.opensrc@micron.com>
Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-2-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/cxl/cxl_device.h | 4 ++
include/hw/cxl/cxl_mailbox.h | 1 +
hw/cxl/cxl-device-utils.c | 14 ++++---
hw/cxl/cxl-mailbox-utils.c | 72 ++++++++++++++++++++++++++++++++----
hw/mem/cxl_type3.c | 8 +++-
5 files changed, 86 insertions(+), 13 deletions(-)
diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
index 3a0ee7e8e7..d21695507f 100644
--- a/include/hw/cxl/cxl_device.h
+++ b/include/hw/cxl/cxl_device.h
@@ -176,10 +176,12 @@ typedef struct CXLCCI {
uint16_t opcode;
uint16_t complete_pct;
uint16_t ret_code; /* Current value of retcode */
+ bool aborted;
uint64_t starttime;
/* set by each bg cmd, cleared by the bg_timer when complete */
uint64_t runtime;
QEMUTimer *timer;
+ QemuMutex lock; /* serializes mbox abort vs timer cb */
} bg;
/* firmware update */
@@ -201,6 +203,7 @@ typedef struct CXLCCI {
DeviceState *d;
/* Pointer to the device hosting the protocol conversion */
DeviceState *intf;
+ bool initialized;
} CXLCCI;
typedef struct cxl_device_state {
@@ -316,6 +319,7 @@ void cxl_initialize_mailbox_t3(CXLCCI *cci, DeviceState *d, size_t payload_max);
void cxl_initialize_mailbox_swcci(CXLCCI *cci, DeviceState *intf,
DeviceState *d, size_t payload_max);
void cxl_init_cci(CXLCCI *cci, size_t payload_max);
+void cxl_destroy_cci(CXLCCI *cci);
void cxl_add_cci_commands(CXLCCI *cci, const struct cxl_cmd (*cxl_cmd_set)[256],
size_t payload_max);
int cxl_process_cci_message(CXLCCI *cci, uint8_t set, uint8_t cmd,
diff --git a/include/hw/cxl/cxl_mailbox.h b/include/hw/cxl/cxl_mailbox.h
index beb048052e..9008402d1c 100644
--- a/include/hw/cxl/cxl_mailbox.h
+++ b/include/hw/cxl/cxl_mailbox.h
@@ -14,5 +14,6 @@
#define CXL_MBOX_IMMEDIATE_LOG_CHANGE (1 << 4)
#define CXL_MBOX_SECURITY_STATE_CHANGE (1 << 5)
#define CXL_MBOX_BACKGROUND_OPERATION (1 << 6)
+#define CXL_MBOX_BACKGROUND_OPERATION_ABORT (1 << 7)
#endif
diff --git a/hw/cxl/cxl-device-utils.c b/hw/cxl/cxl-device-utils.c
index 52ad1e4c3f..e150d74457 100644
--- a/hw/cxl/cxl-device-utils.c
+++ b/hw/cxl/cxl-device-utils.c
@@ -95,11 +95,15 @@ static uint64_t mailbox_reg_read(void *opaque, hwaddr offset, unsigned size)
}
if (offset == A_CXL_DEV_MAILBOX_STS) {
uint64_t status_reg = cxl_dstate->mbox_reg_state64[offset / size];
- if (cci->bg.complete_pct) {
- status_reg = FIELD_DP64(status_reg, CXL_DEV_MAILBOX_STS, BG_OP,
- 0);
- cxl_dstate->mbox_reg_state64[offset / size] = status_reg;
- }
+ int bgop;
+
+ qemu_mutex_lock(&cci->bg.lock);
+ bgop = !(cci->bg.complete_pct == 100 || cci->bg.aborted);
+
+ status_reg = FIELD_DP64(status_reg, CXL_DEV_MAILBOX_STS, BG_OP,
+ bgop);
+ cxl_dstate->mbox_reg_state64[offset / size] = status_reg;
+ qemu_mutex_unlock(&cci->bg.lock);
}
return cxl_dstate->mbox_reg_state64[offset / size];
default:
diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
index 516c01d840..4401f446d9 100644
--- a/hw/cxl/cxl-mailbox-utils.c
+++ b/hw/cxl/cxl-mailbox-utils.c
@@ -56,6 +56,7 @@ enum {
INFOSTAT = 0x00,
#define IS_IDENTIFY 0x1
#define BACKGROUND_OPERATION_STATUS 0x2
+ #define BACKGROUND_OPERATION_ABORT 0x5
EVENTS = 0x01,
#define GET_RECORDS 0x0
#define CLEAR_RECORDS 0x1
@@ -636,6 +637,41 @@ static CXLRetCode cmd_infostat_bg_op_sts(const struct cxl_cmd *cmd,
return CXL_MBOX_SUCCESS;
}
+/*
+ * CXL r3.1 Section 8.2.9.1.5:
+ * Request Abort Background Operation (Opcode 0005h)
+ */
+static CXLRetCode cmd_infostat_bg_op_abort(const struct cxl_cmd *cmd,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ CXLCCI *cci)
+{
+ int bg_set = cci->bg.opcode >> 8;
+ int bg_cmd = cci->bg.opcode & 0xff;
+ const struct cxl_cmd *bg_c = &cci->cxl_cmd_set[bg_set][bg_cmd];
+
+ if (!(bg_c->effect & CXL_MBOX_BACKGROUND_OPERATION_ABORT)) {
+ return CXL_MBOX_REQUEST_ABORT_NOTSUP;
+ }
+
+ qemu_mutex_lock(&cci->bg.lock);
+ if (cci->bg.runtime) {
+ /* operation is near complete, let it finish */
+ if (cci->bg.complete_pct < 85) {
+ timer_del(cci->bg.timer);
+ cci->bg.ret_code = CXL_MBOX_ABORTED;
+ cci->bg.starttime = 0;
+ cci->bg.runtime = 0;
+ cci->bg.aborted = true;
+ }
+ }
+ qemu_mutex_unlock(&cci->bg.lock);
+
+ return CXL_MBOX_SUCCESS;
+}
+
#define CXL_FW_SLOTS 2
#define CXL_FW_SIZE 0x02000000 /* 32 mb */
@@ -2715,6 +2751,8 @@ static CXLRetCode cmd_dcd_release_dyn_cap(const struct cxl_cmd *cmd,
}
static const struct cxl_cmd cxl_cmd_set[256][256] = {
+ [INFOSTAT][BACKGROUND_OPERATION_ABORT] = { "BACKGROUND_OPERATION_ABORT",
+ cmd_infostat_bg_op_abort, 0, 0 },
[EVENTS][GET_RECORDS] = { "EVENTS_GET_RECORDS",
cmd_events_get_records, 1, 0 },
[EVENTS][CLEAR_RECORDS] = { "EVENTS_CLEAR_RECORDS",
@@ -2727,9 +2765,11 @@ static const struct cxl_cmd cxl_cmd_set[256][256] = {
[FIRMWARE_UPDATE][GET_INFO] = { "FIRMWARE_UPDATE_GET_INFO",
cmd_firmware_update_get_info, 0, 0 },
[FIRMWARE_UPDATE][TRANSFER] = { "FIRMWARE_UPDATE_TRANSFER",
- cmd_firmware_update_transfer, ~0, CXL_MBOX_BACKGROUND_OPERATION },
+ cmd_firmware_update_transfer, ~0,
+ CXL_MBOX_BACKGROUND_OPERATION | CXL_MBOX_BACKGROUND_OPERATION_ABORT },
[FIRMWARE_UPDATE][ACTIVATE] = { "FIRMWARE_UPDATE_ACTIVATE",
- cmd_firmware_update_activate, 2, CXL_MBOX_BACKGROUND_OPERATION },
+ cmd_firmware_update_activate, 2,
+ CXL_MBOX_BACKGROUND_OPERATION | CXL_MBOX_BACKGROUND_OPERATION_ABORT },
[TIMESTAMP][GET] = { "TIMESTAMP_GET", cmd_timestamp_get, 0, 0 },
[TIMESTAMP][SET] = { "TIMESTAMP_SET", cmd_timestamp_set,
8, CXL_MBOX_IMMEDIATE_POLICY_CHANGE },
@@ -2758,7 +2798,8 @@ static const struct cxl_cmd cxl_cmd_set[256][256] = {
[SANITIZE][OVERWRITE] = { "SANITIZE_OVERWRITE", cmd_sanitize_overwrite, 0,
(CXL_MBOX_IMMEDIATE_DATA_CHANGE |
CXL_MBOX_SECURITY_STATE_CHANGE |
- CXL_MBOX_BACKGROUND_OPERATION)},
+ CXL_MBOX_BACKGROUND_OPERATION |
+ CXL_MBOX_BACKGROUND_OPERATION_ABORT)},
[PERSISTENT_MEM][GET_SECURITY_STATE] = { "GET_SECURITY_STATE",
cmd_get_security_state, 0, 0 },
[MEDIA_AND_POISON][GET_POISON_LIST] = { "MEDIA_AND_POISON_GET_POISON_LIST",
@@ -2771,7 +2812,8 @@ static const struct cxl_cmd cxl_cmd_set[256][256] = {
"MEDIA_AND_POISON_GET_SCAN_MEDIA_CAPABILITIES",
cmd_media_get_scan_media_capabilities, 16, 0 },
[MEDIA_AND_POISON][SCAN_MEDIA] = { "MEDIA_AND_POISON_SCAN_MEDIA",
- cmd_media_scan_media, 17, CXL_MBOX_BACKGROUND_OPERATION },
+ cmd_media_scan_media, 17,
+ (CXL_MBOX_BACKGROUND_OPERATION | CXL_MBOX_BACKGROUND_OPERATION_ABORT)},
[MEDIA_AND_POISON][GET_SCAN_MEDIA_RESULTS] = {
"MEDIA_AND_POISON_GET_SCAN_MEDIA_RESULTS",
cmd_media_get_scan_media_results, 0, 0 },
@@ -2795,6 +2837,8 @@ static const struct cxl_cmd cxl_cmd_set_sw[256][256] = {
[INFOSTAT][IS_IDENTIFY] = { "IDENTIFY", cmd_infostat_identify, 0, 0 },
[INFOSTAT][BACKGROUND_OPERATION_STATUS] = { "BACKGROUND_OPERATION_STATUS",
cmd_infostat_bg_op_sts, 0, 0 },
+ [INFOSTAT][BACKGROUND_OPERATION_ABORT] = { "BACKGROUND_OPERATION_ABORT",
+ cmd_infostat_bg_op_abort, 0, 0 },
[TIMESTAMP][GET] = { "TIMESTAMP_GET", cmd_timestamp_get, 0, 0 },
[TIMESTAMP][SET] = { "TIMESTAMP_SET", cmd_timestamp_set, 8,
CXL_MBOX_IMMEDIATE_POLICY_CHANGE },
@@ -2881,6 +2925,7 @@ int cxl_process_cci_message(CXLCCI *cci, uint8_t set, uint8_t cmd,
cci->bg.opcode = (set << 8) | cmd;
cci->bg.complete_pct = 0;
+ cci->bg.aborted = false;
cci->bg.ret_code = 0;
now = qemu_clock_get_ms(QEMU_CLOCK_VIRTUAL);
@@ -2894,10 +2939,12 @@ int cxl_process_cci_message(CXLCCI *cci, uint8_t set, uint8_t cmd,
static void bg_timercb(void *opaque)
{
CXLCCI *cci = opaque;
- uint64_t now = qemu_clock_get_ms(QEMU_CLOCK_VIRTUAL);
- uint64_t total_time = cci->bg.starttime + cci->bg.runtime;
+ uint64_t now, total_time;
- assert(cci->bg.runtime > 0);
+ qemu_mutex_lock(&cci->bg.lock);
+
+ now = qemu_clock_get_ms(QEMU_CLOCK_VIRTUAL);
+ total_time = cci->bg.starttime + cci->bg.runtime;
if (now >= total_time) { /* we are done */
uint16_t ret = CXL_MBOX_SUCCESS;
@@ -2950,6 +2997,8 @@ static void bg_timercb(void *opaque)
msi_notify(pdev, cxl_dstate->mbox_msi_n);
}
}
+
+ qemu_mutex_unlock(&cci->bg.lock);
}
static void cxl_rebuild_cel(CXLCCI *cci)
@@ -2978,12 +3027,21 @@ void cxl_init_cci(CXLCCI *cci, size_t payload_max)
cci->bg.complete_pct = 0;
cci->bg.starttime = 0;
cci->bg.runtime = 0;
+ cci->bg.aborted = false;
cci->bg.timer = timer_new_ms(QEMU_CLOCK_VIRTUAL,
bg_timercb, cci);
+ qemu_mutex_init(&cci->bg.lock);
memset(&cci->fw, 0, sizeof(cci->fw));
cci->fw.active_slot = 1;
cci->fw.slot[cci->fw.active_slot - 1] = true;
+ cci->initialized = true;
+}
+
+void cxl_destroy_cci(CXLCCI *cci)
+{
+ qemu_mutex_destroy(&cci->bg.lock);
+ cci->initialized = false;
}
static void cxl_copy_cci_commands(CXLCCI *cci, const struct cxl_cmd (*cxl_cmds)[256])
diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
index bba923f8ea..aacd078118 100644
--- a/hw/mem/cxl_type3.c
+++ b/hw/mem/cxl_type3.c
@@ -969,6 +969,7 @@ static void ct3_exit(PCIDevice *pci_dev)
cxl_doe_cdat_release(cxl_cstate);
msix_uninit_exclusive_bar(pci_dev);
g_free(regs->special_ops);
+ cxl_destroy_cci(&ct3d->cci);
if (ct3d->dc.host_dc) {
cxl_destroy_dc_regions(ct3d);
address_space_destroy(&ct3d->dc.host_dc_as);
@@ -1224,12 +1225,17 @@ static void ct3d_reset(DeviceState *dev)
* Bring up an endpoint to target with MCTP over VDM.
* This device is emulating an MLD with single LD for now.
*/
+ if (ct3d->vdm_fm_owned_ld_mctp_cci.initialized) {
+ cxl_destroy_cci(&ct3d->vdm_fm_owned_ld_mctp_cci);
+ }
cxl_initialize_t3_fm_owned_ld_mctpcci(&ct3d->vdm_fm_owned_ld_mctp_cci,
DEVICE(ct3d), DEVICE(ct3d),
512); /* Max payload made up */
+ if (ct3d->ld0_cci.initialized) {
+ cxl_destroy_cci(&ct3d->ld0_cci);
+ }
cxl_initialize_t3_ld_cci(&ct3d->ld0_cci, DEVICE(ct3d), DEVICE(ct3d),
512); /* Max payload made up */
-
}
static const Property ct3_props[] = {
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 02/27] hw/cxl: Support get/set mctp response payload size
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 01/27] hw/cxl: Support aborting background commands Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 03/27] hw/cxl/cxl-mailbox-utils: Add support for Media operations discovery commands cxl r3.2 (8.2.10.9.5.3) Michael S. Tsirkin
` (25 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Davidlohr Bueso, Fan Ni, Jonathan Cameron
From: Davidlohr Bueso <dave@stgolabs.net>
Add Get/Set Response Message Limit commands.
Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
Reviewed-by: Fan Ni <fan.ni@samsung.com>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-3-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/cxl/cxl-mailbox-utils.c | 58 ++++++++++++++++++++++++++++++++++++--
1 file changed, 56 insertions(+), 2 deletions(-)
diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
index 4401f446d9..bd25df033a 100644
--- a/hw/cxl/cxl-mailbox-utils.c
+++ b/hw/cxl/cxl-mailbox-utils.c
@@ -7,6 +7,8 @@
* COPYING file in the top-level directory.
*/
+#include <math.h>
+
#include "qemu/osdep.h"
#include "hw/pci/msi.h"
#include "hw/pci/msix.h"
@@ -56,6 +58,8 @@ enum {
INFOSTAT = 0x00,
#define IS_IDENTIFY 0x1
#define BACKGROUND_OPERATION_STATUS 0x2
+ #define GET_RESPONSE_MSG_LIMIT 0x3
+ #define SET_RESPONSE_MSG_LIMIT 0x4
#define BACKGROUND_OPERATION_ABORT 0x5
EVENTS = 0x01,
#define GET_RECORDS 0x0
@@ -413,12 +417,58 @@ static CXLRetCode cmd_infostat_identify(const struct cxl_cmd *cmd,
is_identify->component_type = 0x3; /* Type 3 */
}
- /* TODO: Allow this to vary across different CCIs */
- is_identify->max_message_size = 9; /* 512 bytes - MCTP_CXL_MAILBOX_BYTES */
+ is_identify->max_message_size = (uint8_t)log2(cci->payload_max);
*len_out = sizeof(*is_identify);
return CXL_MBOX_SUCCESS;
}
+/* CXL r3.1 section 8.2.9.1.3: Get Response Message Limit (Opcode 0003h) */
+static CXLRetCode cmd_get_response_msg_limit(const struct cxl_cmd *cmd,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ CXLCCI *cci)
+{
+ struct {
+ uint8_t rsp_limit;
+ } QEMU_PACKED *get_rsp_msg_limit = (void *)payload_out;
+ QEMU_BUILD_BUG_ON(sizeof(*get_rsp_msg_limit) != 1);
+
+ get_rsp_msg_limit->rsp_limit = (uint8_t)log2(cci->payload_max);
+
+ *len_out = sizeof(*get_rsp_msg_limit);
+ return CXL_MBOX_SUCCESS;
+}
+
+/* CXL r3.1 section 8.2.9.1.4: Set Response Message Limit (Opcode 0004h) */
+static CXLRetCode cmd_set_response_msg_limit(const struct cxl_cmd *cmd,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ CXLCCI *cci)
+{
+ struct {
+ uint8_t rsp_limit;
+ } QEMU_PACKED *in = (void *)payload_in;
+ QEMU_BUILD_BUG_ON(sizeof(*in) != 1);
+ struct {
+ uint8_t rsp_limit;
+ } QEMU_PACKED *out = (void *)payload_out;
+ QEMU_BUILD_BUG_ON(sizeof(*out) != 1);
+
+ if (in->rsp_limit < 8 || in->rsp_limit > 10) {
+ return CXL_MBOX_INVALID_INPUT;
+ }
+
+ cci->payload_max = 1 << in->rsp_limit;
+ out->rsp_limit = in->rsp_limit;
+
+ *len_out = sizeof(*out);
+ return CXL_MBOX_SUCCESS;
+}
+
static void cxl_set_dsp_active_bm(PCIBus *b, PCIDevice *d,
void *private)
{
@@ -3105,6 +3155,10 @@ void cxl_initialize_t3_ld_cci(CXLCCI *cci, DeviceState *d, DeviceState *intf,
static const struct cxl_cmd cxl_cmd_set_t3_fm_owned_ld_mctp[256][256] = {
[INFOSTAT][IS_IDENTIFY] = { "IDENTIFY", cmd_infostat_identify, 0, 0},
+ [INFOSTAT][GET_RESPONSE_MSG_LIMIT] = { "GET_RESPONSE_MSG_LIMIT",
+ cmd_get_response_msg_limit, 0, 0 },
+ [INFOSTAT][SET_RESPONSE_MSG_LIMIT] = { "SET_RESPONSE_MSG_LIMIT",
+ cmd_set_response_msg_limit, 1, 0 },
[LOGS][GET_SUPPORTED] = { "LOGS_GET_SUPPORTED", cmd_logs_get_supported, 0,
0 },
[LOGS][GET_LOG] = { "LOGS_GET_LOG", cmd_logs_get_log, 0x18, 0 },
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 03/27] hw/cxl/cxl-mailbox-utils: Add support for Media operations discovery commands cxl r3.2 (8.2.10.9.5.3)
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 01/27] hw/cxl: Support aborting background commands Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 02/27] hw/cxl: Support get/set mctp response payload size Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-07-10 13:26 ` Peter Maydell
2025-05-14 11:50 ` [PULL 04/27] hw/cxl: factor out calculation of sanitize duration from cmd_santize_overwrite Michael S. Tsirkin
` (24 subsequent siblings)
27 siblings, 1 reply; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Vinayak Holikatti, Jonathan Cameron, Fan Ni
From: Vinayak Holikatti <vinayak.kh@samsung.com>
CXL spec 3.2 section 8.2.10.9.5.3 describes media operations commands.
CXL devices supports media operations discovery command.
Signed-off-by: Vinayak Holikatti <vinayak.kh@samsung.com>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-4-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/cxl/cxl-mailbox-utils.c | 125 +++++++++++++++++++++++++++++++++++++
1 file changed, 125 insertions(+)
diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
index bd25df033a..79b35d1405 100644
--- a/hw/cxl/cxl-mailbox-utils.c
+++ b/hw/cxl/cxl-mailbox-utils.c
@@ -89,6 +89,7 @@ enum {
SANITIZE = 0x44,
#define OVERWRITE 0x0
#define SECURE_ERASE 0x1
+ #define MEDIA_OPERATIONS 0x2
PERSISTENT_MEM = 0x45,
#define GET_SECURITY_STATE 0x0
MEDIA_AND_POISON = 0x43,
@@ -1705,6 +1706,126 @@ static CXLRetCode cmd_sanitize_overwrite(const struct cxl_cmd *cmd,
return CXL_MBOX_BG_STARTED;
}
+enum {
+ MEDIA_OP_CLASS_GENERAL = 0x0,
+ #define MEDIA_OP_GEN_SUBC_DISCOVERY 0x0
+ MEDIA_OP_CLASS_SANITIZE = 0x1,
+ #define MEDIA_OP_SAN_SUBC_SANITIZE 0x0
+ #define MEDIA_OP_SAN_SUBC_ZERO 0x1
+};
+
+struct media_op_supported_list_entry {
+ uint8_t media_op_class;
+ uint8_t media_op_subclass;
+};
+
+struct media_op_discovery_out_pl {
+ uint64_t dpa_range_granularity;
+ uint16_t total_supported_operations;
+ uint16_t num_of_supported_operations;
+ struct media_op_supported_list_entry entry[];
+} QEMU_PACKED;
+
+static const struct media_op_supported_list_entry media_op_matrix[] = {
+ { MEDIA_OP_CLASS_GENERAL, MEDIA_OP_GEN_SUBC_DISCOVERY },
+ { MEDIA_OP_CLASS_SANITIZE, MEDIA_OP_SAN_SUBC_SANITIZE },
+ { MEDIA_OP_CLASS_SANITIZE, MEDIA_OP_SAN_SUBC_ZERO },
+};
+
+static CXLRetCode media_operations_discovery(uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out)
+{
+ struct {
+ uint8_t media_operation_class;
+ uint8_t media_operation_subclass;
+ uint8_t rsvd[2];
+ uint32_t dpa_range_count;
+ struct {
+ uint16_t start_index;
+ uint16_t num_ops;
+ } discovery_osa;
+ } QEMU_PACKED *media_op_in_disc_pl = (void *)payload_in;
+ struct media_op_discovery_out_pl *media_out_pl =
+ (struct media_op_discovery_out_pl *)payload_out;
+ int num_ops, start_index, i;
+ int count = 0;
+
+ if (len_in < sizeof(*media_op_in_disc_pl)) {
+ return CXL_MBOX_INVALID_PAYLOAD_LENGTH;
+ }
+
+ num_ops = media_op_in_disc_pl->discovery_osa.num_ops;
+ start_index = media_op_in_disc_pl->discovery_osa.start_index;
+
+ /*
+ * As per spec CXL r3.2 8.2.10.9.5.3 dpa_range_count should be zero and
+ * start index should not exceed the total number of entries for discovery
+ * sub class command.
+ */
+ if (media_op_in_disc_pl->dpa_range_count ||
+ start_index > ARRAY_SIZE(media_op_matrix)) {
+ return CXL_MBOX_INVALID_INPUT;
+ }
+
+ media_out_pl->dpa_range_granularity = CXL_CACHE_LINE_SIZE;
+ media_out_pl->total_supported_operations =
+ ARRAY_SIZE(media_op_matrix);
+ if (num_ops > 0) {
+ for (i = start_index; i < start_index + num_ops; i++) {
+ media_out_pl->entry[count].media_op_class =
+ media_op_matrix[i].media_op_class;
+ media_out_pl->entry[count].media_op_subclass =
+ media_op_matrix[i].media_op_subclass;
+ count++;
+ if (count == num_ops) {
+ break;
+ }
+ }
+ }
+
+ media_out_pl->num_of_supported_operations = count;
+ *len_out = sizeof(*media_out_pl) + count * sizeof(*media_out_pl->entry);
+ return CXL_MBOX_SUCCESS;
+}
+
+static CXLRetCode cmd_media_operations(const struct cxl_cmd *cmd,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ CXLCCI *cci)
+{
+ struct {
+ uint8_t media_operation_class;
+ uint8_t media_operation_subclass;
+ uint8_t rsvd[2];
+ uint32_t dpa_range_count;
+ } QEMU_PACKED *media_op_in_common_pl = (void *)payload_in;
+ uint8_t media_op_cl = 0;
+ uint8_t media_op_subclass = 0;
+
+ if (len_in < sizeof(*media_op_in_common_pl)) {
+ return CXL_MBOX_INVALID_PAYLOAD_LENGTH;
+ }
+
+ media_op_cl = media_op_in_common_pl->media_operation_class;
+ media_op_subclass = media_op_in_common_pl->media_operation_subclass;
+
+ switch (media_op_cl) {
+ case MEDIA_OP_CLASS_GENERAL:
+ if (media_op_subclass != MEDIA_OP_GEN_SUBC_DISCOVERY) {
+ return CXL_MBOX_UNSUPPORTED;
+ }
+
+ return media_operations_discovery(payload_in, len_in, payload_out,
+ len_out);
+ default:
+ return CXL_MBOX_UNSUPPORTED;
+ }
+}
+
static CXLRetCode cmd_get_security_state(const struct cxl_cmd *cmd,
uint8_t *payload_in,
size_t len_in,
@@ -2850,6 +2971,10 @@ static const struct cxl_cmd cxl_cmd_set[256][256] = {
CXL_MBOX_SECURITY_STATE_CHANGE |
CXL_MBOX_BACKGROUND_OPERATION |
CXL_MBOX_BACKGROUND_OPERATION_ABORT)},
+ [SANITIZE][MEDIA_OPERATIONS] = { "MEDIA_OPERATIONS", cmd_media_operations,
+ ~0,
+ (CXL_MBOX_IMMEDIATE_DATA_CHANGE |
+ CXL_MBOX_BACKGROUND_OPERATION)},
[PERSISTENT_MEM][GET_SECURITY_STATE] = { "GET_SECURITY_STATE",
cmd_get_security_state, 0, 0 },
[MEDIA_AND_POISON][GET_POISON_LIST] = { "MEDIA_AND_POISON_GET_POISON_LIST",
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PULL 03/27] hw/cxl/cxl-mailbox-utils: Add support for Media operations discovery commands cxl r3.2 (8.2.10.9.5.3)
2025-05-14 11:50 ` [PULL 03/27] hw/cxl/cxl-mailbox-utils: Add support for Media operations discovery commands cxl r3.2 (8.2.10.9.5.3) Michael S. Tsirkin
@ 2025-07-10 13:26 ` Peter Maydell
0 siblings, 0 replies; 31+ messages in thread
From: Peter Maydell @ 2025-07-10 13:26 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: qemu-devel, Vinayak Holikatti, Jonathan Cameron, Fan Ni
On Wed, 14 May 2025 at 12:50, Michael S. Tsirkin <mst@redhat.com> wrote:
>
> From: Vinayak Holikatti <vinayak.kh@samsung.com>
>
> CXL spec 3.2 section 8.2.10.9.5.3 describes media operations commands.
> CXL devices supports media operations discovery command.
>
> Signed-off-by: Vinayak Holikatti <vinayak.kh@samsung.com>
> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> Message-Id: <20250305092501.191929-4-Jonathan.Cameron@huawei.com>
> Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> ---
> hw/cxl/cxl-mailbox-utils.c | 125 +++++++++++++++++++++++++++++++++++++
> 1 file changed, 125 insertions(+)
> +static CXLRetCode media_operations_discovery(uint8_t *payload_in,
> + size_t len_in,
> + uint8_t *payload_out,
> + size_t *len_out)
> +{
> + struct {
> + uint8_t media_operation_class;
> + uint8_t media_operation_subclass;
> + uint8_t rsvd[2];
> + uint32_t dpa_range_count;
> + struct {
> + uint16_t start_index;
> + uint16_t num_ops;
> + } discovery_osa;
> + } QEMU_PACKED *media_op_in_disc_pl = (void *)payload_in;
> + struct media_op_discovery_out_pl *media_out_pl =
> + (struct media_op_discovery_out_pl *)payload_out;
> + int num_ops, start_index, i;
> + int count = 0;
> +
> + if (len_in < sizeof(*media_op_in_disc_pl)) {
> + return CXL_MBOX_INVALID_PAYLOAD_LENGTH;
> + }
> +
> + num_ops = media_op_in_disc_pl->discovery_osa.num_ops;
> + start_index = media_op_in_disc_pl->discovery_osa.start_index;
> +
> + /*
> + * As per spec CXL r3.2 8.2.10.9.5.3 dpa_range_count should be zero and
> + * start index should not exceed the total number of entries for discovery
> + * sub class command.
> + */
> + if (media_op_in_disc_pl->dpa_range_count ||
> + start_index > ARRAY_SIZE(media_op_matrix)) {
Coverity thinks this bounds check is wrong (CID 1610091):
we allow start_index equal to the ARRAY_SIZE(media_op_matrix),
which means that in the loop below we will index off the
end of the array.
Don't we also need to be checking (start_index + num_ops)
against the array size bounds, not just start_index ?
> + return CXL_MBOX_INVALID_INPUT;
> + }
> +
> + media_out_pl->dpa_range_granularity = CXL_CACHE_LINE_SIZE;
> + media_out_pl->total_supported_operations =
> + ARRAY_SIZE(media_op_matrix);
> + if (num_ops > 0) {
> + for (i = start_index; i < start_index + num_ops; i++) {
> + media_out_pl->entry[count].media_op_class =
> + media_op_matrix[i].media_op_class;
> + media_out_pl->entry[count].media_op_subclass =
> + media_op_matrix[i].media_op_subclass;
> + count++;
> + if (count == num_ops) {
> + break;
> + }
> + }
> + }
> +
> + media_out_pl->num_of_supported_operations = count;
> + *len_out = sizeof(*media_out_pl) + count * sizeof(*media_out_pl->entry);
> + return CXL_MBOX_SUCCESS;
> +}
The functions in this patch also look like they aren't correctly
handling the "guest and host endianness differ" case.
thanks
-- PMM
^ permalink raw reply [flat|nested] 31+ messages in thread
* [PULL 04/27] hw/cxl: factor out calculation of sanitize duration from cmd_santize_overwrite
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (2 preceding siblings ...)
2025-05-14 11:50 ` [PULL 03/27] hw/cxl/cxl-mailbox-utils: Add support for Media operations discovery commands cxl r3.2 (8.2.10.9.5.3) Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 05/27] hw/cxl/cxl-mailbox-utils: Media operations Sanitize and Write Zeros commands CXL r3.2(8.2.10.9.5.3) Michael S. Tsirkin
` (23 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Vinayak Holikatti, Davidlohr Bueso,
Jonathan Cameron, Fan Ni
From: Vinayak Holikatti <vinayak.kh@samsung.com>
Move the code of calculation of sanitize duration into function
for usability by other sanitize routines
Estimate times based on:
https://pmem.io/documents/NVDIMM_DSM_Interface-V1.8.pdf
Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
Signed-off-by: Vinayak Holikatti <vinayak.kh@samsung.com>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-5-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/cxl/cxl-mailbox-utils.c | 63 ++++++++++++++++++++++----------------
1 file changed, 36 insertions(+), 27 deletions(-)
diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
index 79b35d1405..9f9d475678 100644
--- a/hw/cxl/cxl-mailbox-utils.c
+++ b/hw/cxl/cxl-mailbox-utils.c
@@ -1640,6 +1640,41 @@ static void __do_sanitization(CXLType3Dev *ct3d)
cxl_discard_all_event_records(&ct3d->cxl_dstate);
}
+static int get_sanitize_duration(uint64_t total_mem)
+{
+ int secs = 0;
+
+ if (total_mem <= 512) {
+ secs = 4;
+ } else if (total_mem <= 1024) {
+ secs = 8;
+ } else if (total_mem <= 2 * 1024) {
+ secs = 15;
+ } else if (total_mem <= 4 * 1024) {
+ secs = 30;
+ } else if (total_mem <= 8 * 1024) {
+ secs = 60;
+ } else if (total_mem <= 16 * 1024) {
+ secs = 2 * 60;
+ } else if (total_mem <= 32 * 1024) {
+ secs = 4 * 60;
+ } else if (total_mem <= 64 * 1024) {
+ secs = 8 * 60;
+ } else if (total_mem <= 128 * 1024) {
+ secs = 15 * 60;
+ } else if (total_mem <= 256 * 1024) {
+ secs = 30 * 60;
+ } else if (total_mem <= 512 * 1024) {
+ secs = 60 * 60;
+ } else if (total_mem <= 1024 * 1024) {
+ secs = 120 * 60;
+ } else {
+ secs = 240 * 60; /* max 4 hrs */
+ }
+
+ return secs;
+}
+
/*
* CXL r3.1 Section 8.2.9.9.5.1: Sanitize (Opcode 4400h)
*
@@ -1668,33 +1703,7 @@ static CXLRetCode cmd_sanitize_overwrite(const struct cxl_cmd *cmd,
int secs;
total_mem = (ct3d->cxl_dstate.vmem_size + ct3d->cxl_dstate.pmem_size) >> 20;
- if (total_mem <= 512) {
- secs = 4;
- } else if (total_mem <= 1024) {
- secs = 8;
- } else if (total_mem <= 2 * 1024) {
- secs = 15;
- } else if (total_mem <= 4 * 1024) {
- secs = 30;
- } else if (total_mem <= 8 * 1024) {
- secs = 60;
- } else if (total_mem <= 16 * 1024) {
- secs = 2 * 60;
- } else if (total_mem <= 32 * 1024) {
- secs = 4 * 60;
- } else if (total_mem <= 64 * 1024) {
- secs = 8 * 60;
- } else if (total_mem <= 128 * 1024) {
- secs = 15 * 60;
- } else if (total_mem <= 256 * 1024) {
- secs = 30 * 60;
- } else if (total_mem <= 512 * 1024) {
- secs = 60 * 60;
- } else if (total_mem <= 1024 * 1024) {
- secs = 120 * 60;
- } else {
- secs = 240 * 60; /* max 4 hrs */
- }
+ secs = get_sanitize_duration(total_mem);
/* EBUSY other bg cmds as of now */
cci->bg.runtime = secs * 1000UL;
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 05/27] hw/cxl/cxl-mailbox-utils: Media operations Sanitize and Write Zeros commands CXL r3.2(8.2.10.9.5.3)
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (3 preceding siblings ...)
2025-05-14 11:50 ` [PULL 04/27] hw/cxl: factor out calculation of sanitize duration from cmd_santize_overwrite Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-07-10 13:23 ` Peter Maydell
2025-05-14 11:50 ` [PULL 06/27] hw/cxl/cxl-mailbox-utils: CXL CCI Get/Set alert config commands Michael S. Tsirkin
` (22 subsequent siblings)
27 siblings, 1 reply; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Vinayak Holikatti, Jonathan Cameron, Fan Ni
From: Vinayak Holikatti <vinayak.kh@samsung.com>
CXL spec 3.2 section 8.2.10.9.5.3 describes media operations commands.
CXL devices supports media operations Sanitize and Write zero command.
Signed-off-by: Vinayak Holikatti <vinayak.kh@samsung.com>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-6-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/cxl/cxl_device.h | 4 +
hw/cxl/cxl-mailbox-utils.c | 204 ++++++++++++++++++++++++++++++++++++
2 files changed, 208 insertions(+)
diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
index d21695507f..3ec7be3809 100644
--- a/include/hw/cxl/cxl_device.h
+++ b/include/hw/cxl/cxl_device.h
@@ -540,6 +540,8 @@ typedef struct CXLSetFeatureInfo {
size_t data_size;
} CXLSetFeatureInfo;
+struct CXLSanitizeInfo;
+
struct CXLType3Dev {
/* Private */
PCIDevice parent_obj;
@@ -606,6 +608,8 @@ struct CXLType3Dev {
uint8_t num_regions; /* 0-8 regions */
CXLDCRegion regions[DCD_MAX_NUM_REGION];
} dc;
+
+ struct CXLSanitizeInfo *media_op_sanitize;
};
#define TYPE_CXL_TYPE3 "cxl-type3"
diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
index 9f9d475678..2c6db70e5f 100644
--- a/hw/cxl/cxl-mailbox-utils.c
+++ b/hw/cxl/cxl-mailbox-utils.c
@@ -1715,6 +1715,131 @@ static CXLRetCode cmd_sanitize_overwrite(const struct cxl_cmd *cmd,
return CXL_MBOX_BG_STARTED;
}
+struct dpa_range_list_entry {
+ uint64_t starting_dpa;
+ uint64_t length;
+} QEMU_PACKED;
+
+struct CXLSanitizeInfo {
+ uint32_t dpa_range_count;
+ uint8_t fill_value;
+ struct dpa_range_list_entry dpa_range_list[];
+} QEMU_PACKED;
+
+static uint64_t get_vmr_size(CXLType3Dev *ct3d, MemoryRegion **vmr)
+{
+ MemoryRegion *mr;
+ if (ct3d->hostvmem) {
+ mr = host_memory_backend_get_memory(ct3d->hostvmem);
+ if (vmr) {
+ *vmr = mr;
+ }
+ return memory_region_size(mr);
+ }
+ return 0;
+}
+
+static uint64_t get_pmr_size(CXLType3Dev *ct3d, MemoryRegion **pmr)
+{
+ MemoryRegion *mr;
+ if (ct3d->hostpmem) {
+ mr = host_memory_backend_get_memory(ct3d->hostpmem);
+ if (pmr) {
+ *pmr = mr;
+ }
+ return memory_region_size(mr);
+ }
+ return 0;
+}
+
+static uint64_t get_dc_size(CXLType3Dev *ct3d, MemoryRegion **dc_mr)
+{
+ MemoryRegion *mr;
+ if (ct3d->dc.host_dc) {
+ mr = host_memory_backend_get_memory(ct3d->dc.host_dc);
+ if (dc_mr) {
+ *dc_mr = mr;
+ }
+ return memory_region_size(mr);
+ }
+ return 0;
+}
+
+static int validate_dpa_addr(CXLType3Dev *ct3d, uint64_t dpa_addr,
+ size_t length)
+{
+ uint64_t vmr_size, pmr_size, dc_size;
+
+ if ((dpa_addr % CXL_CACHE_LINE_SIZE) ||
+ (length % CXL_CACHE_LINE_SIZE) ||
+ (length <= 0)) {
+ return -EINVAL;
+ }
+
+ vmr_size = get_vmr_size(ct3d, NULL);
+ pmr_size = get_pmr_size(ct3d, NULL);
+ dc_size = get_dc_size(ct3d, NULL);
+
+ if (dpa_addr + length > vmr_size + pmr_size + dc_size) {
+ return -EINVAL;
+ }
+
+ if (dpa_addr > vmr_size + pmr_size) {
+ if (!ct3_test_region_block_backed(ct3d, dpa_addr, length)) {
+ return -ENODEV;
+ }
+ }
+
+ return 0;
+}
+
+static int sanitize_range(CXLType3Dev *ct3d, uint64_t dpa_addr, size_t length,
+ uint8_t fill_value)
+{
+
+ uint64_t vmr_size, pmr_size;
+ AddressSpace *as = NULL;
+ MemTxAttrs mem_attrs = {};
+
+ vmr_size = get_vmr_size(ct3d, NULL);
+ pmr_size = get_pmr_size(ct3d, NULL);
+
+ if (dpa_addr < vmr_size) {
+ as = &ct3d->hostvmem_as;
+ } else if (dpa_addr < vmr_size + pmr_size) {
+ as = &ct3d->hostpmem_as;
+ } else {
+ if (!ct3_test_region_block_backed(ct3d, dpa_addr, length)) {
+ return -ENODEV;
+ }
+ as = &ct3d->dc.host_dc_as;
+ }
+
+ return address_space_set(as, dpa_addr, fill_value, length, mem_attrs);
+}
+
+/* Perform the actual device zeroing */
+static void __do_sanitize(CXLType3Dev *ct3d)
+{
+ struct CXLSanitizeInfo *san_info = ct3d->media_op_sanitize;
+ int dpa_range_count = san_info->dpa_range_count;
+ int rc = 0;
+ int i;
+
+ for (i = 0; i < dpa_range_count; i++) {
+ rc = sanitize_range(ct3d, san_info->dpa_range_list[i].starting_dpa,
+ san_info->dpa_range_list[i].length,
+ san_info->fill_value);
+ if (rc) {
+ goto exit;
+ }
+ }
+exit:
+ g_free(ct3d->media_op_sanitize);
+ ct3d->media_op_sanitize = NULL;
+ return;
+}
+
enum {
MEDIA_OP_CLASS_GENERAL = 0x0,
#define MEDIA_OP_GEN_SUBC_DISCOVERY 0x0
@@ -1799,6 +1924,65 @@ static CXLRetCode media_operations_discovery(uint8_t *payload_in,
return CXL_MBOX_SUCCESS;
}
+static CXLRetCode media_operations_sanitize(CXLType3Dev *ct3d,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ uint8_t fill_value,
+ CXLCCI *cci)
+{
+ struct media_operations_sanitize {
+ uint8_t media_operation_class;
+ uint8_t media_operation_subclass;
+ uint8_t rsvd[2];
+ uint32_t dpa_range_count;
+ struct dpa_range_list_entry dpa_range_list[];
+ } QEMU_PACKED *media_op_in_sanitize_pl = (void *)payload_in;
+ uint32_t dpa_range_count = media_op_in_sanitize_pl->dpa_range_count;
+ uint64_t total_mem = 0;
+ size_t dpa_range_list_size;
+ int secs = 0, i;
+
+ if (dpa_range_count == 0) {
+ return CXL_MBOX_SUCCESS;
+ }
+
+ dpa_range_list_size = dpa_range_count * sizeof(struct dpa_range_list_entry);
+ if (len_in < (sizeof(*media_op_in_sanitize_pl) + dpa_range_list_size)) {
+ return CXL_MBOX_INVALID_PAYLOAD_LENGTH;
+ }
+
+ for (i = 0; i < dpa_range_count; i++) {
+ uint64_t start_dpa =
+ media_op_in_sanitize_pl->dpa_range_list[i].starting_dpa;
+ uint64_t length = media_op_in_sanitize_pl->dpa_range_list[i].length;
+
+ if (validate_dpa_addr(ct3d, start_dpa, length)) {
+ return CXL_MBOX_INVALID_INPUT;
+ }
+ total_mem += length;
+ }
+ ct3d->media_op_sanitize = g_malloc0(sizeof(struct CXLSanitizeInfo) +
+ dpa_range_list_size);
+
+ ct3d->media_op_sanitize->dpa_range_count = dpa_range_count;
+ ct3d->media_op_sanitize->fill_value = fill_value;
+ memcpy(ct3d->media_op_sanitize->dpa_range_list,
+ media_op_in_sanitize_pl->dpa_range_list,
+ dpa_range_list_size);
+ secs = get_sanitize_duration(total_mem >> 20);
+
+ /* EBUSY other bg cmds as of now */
+ cci->bg.runtime = secs * 1000UL;
+ *len_out = 0;
+ /*
+ * media op sanitize is targeted so no need to disable media or
+ * clear event logs
+ */
+ return CXL_MBOX_BG_STARTED;
+}
+
static CXLRetCode cmd_media_operations(const struct cxl_cmd *cmd,
uint8_t *payload_in,
size_t len_in,
@@ -1812,6 +1996,7 @@ static CXLRetCode cmd_media_operations(const struct cxl_cmd *cmd,
uint8_t rsvd[2];
uint32_t dpa_range_count;
} QEMU_PACKED *media_op_in_common_pl = (void *)payload_in;
+ CXLType3Dev *ct3d = CXL_TYPE3(cci->d);
uint8_t media_op_cl = 0;
uint8_t media_op_subclass = 0;
@@ -1830,6 +2015,19 @@ static CXLRetCode cmd_media_operations(const struct cxl_cmd *cmd,
return media_operations_discovery(payload_in, len_in, payload_out,
len_out);
+ case MEDIA_OP_CLASS_SANITIZE:
+ switch (media_op_subclass) {
+ case MEDIA_OP_SAN_SUBC_SANITIZE:
+ return media_operations_sanitize(ct3d, payload_in, len_in,
+ payload_out, len_out, 0xF,
+ cci);
+ case MEDIA_OP_SAN_SUBC_ZERO:
+ return media_operations_sanitize(ct3d, payload_in, len_in,
+ payload_out, len_out, 0,
+ cci);
+ default:
+ return CXL_MBOX_UNSUPPORTED;
+ }
default:
return CXL_MBOX_UNSUPPORTED;
}
@@ -3147,6 +3345,12 @@ static void bg_timercb(void *opaque)
cxl_dev_enable_media(&ct3d->cxl_dstate);
}
break;
+ case 0x4402: /* Media Operations sanitize */
+ {
+ CXLType3Dev *ct3d = CXL_TYPE3(cci->d);
+ __do_sanitize(ct3d);
+ }
+ break;
case 0x4304: /* scan media */
{
CXLType3Dev *ct3d = CXL_TYPE3(cci->d);
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PULL 05/27] hw/cxl/cxl-mailbox-utils: Media operations Sanitize and Write Zeros commands CXL r3.2(8.2.10.9.5.3)
2025-05-14 11:50 ` [PULL 05/27] hw/cxl/cxl-mailbox-utils: Media operations Sanitize and Write Zeros commands CXL r3.2(8.2.10.9.5.3) Michael S. Tsirkin
@ 2025-07-10 13:23 ` Peter Maydell
0 siblings, 0 replies; 31+ messages in thread
From: Peter Maydell @ 2025-07-10 13:23 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: qemu-devel, Vinayak Holikatti, Jonathan Cameron, Fan Ni
On Wed, 14 May 2025 at 12:50, Michael S. Tsirkin <mst@redhat.com> wrote:
>
> From: Vinayak Holikatti <vinayak.kh@samsung.com>
>
> CXL spec 3.2 section 8.2.10.9.5.3 describes media operations commands.
> CXL devices supports media operations Sanitize and Write zero command.
>
> Signed-off-by: Vinayak Holikatti <vinayak.kh@samsung.com>
> Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
> Message-Id: <20250305092501.191929-6-Jonathan.Cameron@huawei.com>
> Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> ---
> +static int validate_dpa_addr(CXLType3Dev *ct3d, uint64_t dpa_addr,
> + size_t length)
> +{
> + uint64_t vmr_size, pmr_size, dc_size;
> +
> + if ((dpa_addr % CXL_CACHE_LINE_SIZE) ||
> + (length % CXL_CACHE_LINE_SIZE) ||
> + (length <= 0)) {
> + return -EINVAL;
> + }
> +
> + vmr_size = get_vmr_size(ct3d, NULL);
> + pmr_size = get_pmr_size(ct3d, NULL);
> + dc_size = get_dc_size(ct3d, NULL);
> +
> + if (dpa_addr + length > vmr_size + pmr_size + dc_size) {
Hi; Coverity flagged up a potential issue in this function (CID 1610093)
Partly it is a false positive (it thinks vmr_size etc can
be -1, but they won't I assume ever be memory regions of that
size), but it did make me notice that this address/length
check looks wrong. If the guest can pass us a (dpa_addr, length)
combination that overflows a 64-bit integer then we can
incorrectly pass this length test.
> + return -EINVAL;
> + }
> +
> + if (dpa_addr > vmr_size + pmr_size) {
> + if (!ct3_test_region_block_backed(ct3d, dpa_addr, length)) {
> + return -ENODEV;
> + }
> + }
> +
> + return 0;
> +}
> +static CXLRetCode media_operations_sanitize(CXLType3Dev *ct3d,
> + uint8_t *payload_in,
> + size_t len_in,
> + uint8_t *payload_out,
> + size_t *len_out,
> + uint8_t fill_value,
> + CXLCCI *cci)
> +{
> + struct media_operations_sanitize {
> + uint8_t media_operation_class;
> + uint8_t media_operation_subclass;
> + uint8_t rsvd[2];
> + uint32_t dpa_range_count;
> + struct dpa_range_list_entry dpa_range_list[];
> + } QEMU_PACKED *media_op_in_sanitize_pl = (void *)payload_in;
> + uint32_t dpa_range_count = media_op_in_sanitize_pl->dpa_range_count;
This looks dubious -- a packed struct presumably from the
guest, with a 32-bit value, that we are reading without
doing any handling of host endianness ?
thanks
-- PMM
^ permalink raw reply [flat|nested] 31+ messages in thread
* [PULL 06/27] hw/cxl/cxl-mailbox-utils: CXL CCI Get/Set alert config commands
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (4 preceding siblings ...)
2025-05-14 11:50 ` [PULL 05/27] hw/cxl/cxl-mailbox-utils: Media operations Sanitize and Write Zeros commands CXL r3.2(8.2.10.9.5.3) Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 07/27] docs/cxl: Add serial number for persistent-memdev Michael S. Tsirkin
` (21 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Sweta Kumari, Jonathan Cameron, Fan Ni
From: Sweta Kumari <s5.kumari@samsung.com>
1) get alert configuration(Opcode 4201h)
2) set alert configuration(Opcode 4202h)
Signed-off-by: Sweta Kumari <s5.kumari@samsung.com>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-7-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/cxl/cxl_device.h | 15 ++++++
hw/cxl/cxl-mailbox-utils.c | 105 ++++++++++++++++++++++++++++++++++++
hw/mem/cxl_type3.c | 14 +++++
3 files changed, 134 insertions(+)
diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
index 3ec7be3809..ed6cd50c67 100644
--- a/include/hw/cxl/cxl_device.h
+++ b/include/hw/cxl/cxl_device.h
@@ -542,6 +542,19 @@ typedef struct CXLSetFeatureInfo {
struct CXLSanitizeInfo;
+typedef struct CXLAlertConfig {
+ uint8_t valid_alerts;
+ uint8_t enable_alerts;
+ uint8_t life_used_crit_alert_thresh;
+ uint8_t life_used_warn_thresh;
+ uint16_t over_temp_crit_alert_thresh;
+ uint16_t under_temp_crit_alert_thresh;
+ uint16_t over_temp_warn_thresh;
+ uint16_t under_temp_warn_thresh;
+ uint16_t cor_vmem_err_warn_thresh;
+ uint16_t cor_pmem_err_warn_thresh;
+} QEMU_PACKED CXLAlertConfig;
+
struct CXLType3Dev {
/* Private */
PCIDevice parent_obj;
@@ -563,6 +576,8 @@ struct CXLType3Dev {
CXLCCI vdm_fm_owned_ld_mctp_cci;
CXLCCI ld0_cci;
+ CXLAlertConfig alert_config;
+
/* PCIe link characteristics */
PCIExpLinkSpeed speed;
PCIExpLinkWidth width;
diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c
index 2c6db70e5f..299f232f26 100644
--- a/hw/cxl/cxl-mailbox-utils.c
+++ b/hw/cxl/cxl-mailbox-utils.c
@@ -28,6 +28,11 @@
#define CXL_DC_EVENT_LOG_SIZE 8
#define CXL_NUM_EXTENTS_SUPPORTED 512
#define CXL_NUM_TAGS_SUPPORTED 0
+#define CXL_ALERTS_LIFE_USED_WARN_THRESH (1 << 0)
+#define CXL_ALERTS_OVER_TEMP_WARN_THRESH (1 << 1)
+#define CXL_ALERTS_UNDER_TEMP_WARN_THRESH (1 << 2)
+#define CXL_ALERTS_COR_VMEM_ERR_WARN_THRESH (1 << 3)
+#define CXL_ALERTS_COR_PMEM_ERR_WARN_THRESH (1 << 4)
/*
* How to add a new command, example. The command set FOO, with cmd BAR.
@@ -86,6 +91,9 @@ enum {
#define GET_PARTITION_INFO 0x0
#define GET_LSA 0x2
#define SET_LSA 0x3
+ HEALTH_INFO_ALERTS = 0x42,
+ #define GET_ALERT_CONFIG 0x1
+ #define SET_ALERT_CONFIG 0x2
SANITIZE = 0x44,
#define OVERWRITE 0x0
#define SECURE_ERASE 0x1
@@ -1610,6 +1618,97 @@ static CXLRetCode cmd_ccls_set_lsa(const struct cxl_cmd *cmd,
return CXL_MBOX_SUCCESS;
}
+/* CXL r3.2 Section 8.2.10.9.3.2 Get Alert Configuration (Opcode 4201h) */
+static CXLRetCode cmd_get_alert_config(const struct cxl_cmd *cmd,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ CXLCCI *cci)
+{
+ CXLType3Dev *ct3d = CXL_TYPE3(cci->d);
+ CXLAlertConfig *out = (CXLAlertConfig *)payload_out;
+
+ memcpy(out, &ct3d->alert_config, sizeof(ct3d->alert_config));
+ *len_out = sizeof(ct3d->alert_config);
+
+ return CXL_MBOX_SUCCESS;
+}
+
+/* CXL r3.2 Section 8.2.10.9.3.3 Set Alert Configuration (Opcode 4202h) */
+static CXLRetCode cmd_set_alert_config(const struct cxl_cmd *cmd,
+ uint8_t *payload_in,
+ size_t len_in,
+ uint8_t *payload_out,
+ size_t *len_out,
+ CXLCCI *cci)
+{
+ CXLType3Dev *ct3d = CXL_TYPE3(cci->d);
+ CXLAlertConfig *alert_config = &ct3d->alert_config;
+ struct {
+ uint8_t valid_alert_actions;
+ uint8_t enable_alert_actions;
+ uint8_t life_used_warn_thresh;
+ uint8_t rsvd;
+ uint16_t over_temp_warn_thresh;
+ uint16_t under_temp_warn_thresh;
+ uint16_t cor_vmem_err_warn_thresh;
+ uint16_t cor_pmem_err_warn_thresh;
+ } QEMU_PACKED *in = (void *)payload_in;
+
+ if (in->valid_alert_actions & CXL_ALERTS_LIFE_USED_WARN_THRESH) {
+ /*
+ * CXL r3.2 Table 8-149 The life used warning threshold shall be
+ * less than the life used critical alert value.
+ */
+ if (in->life_used_warn_thresh >=
+ alert_config->life_used_crit_alert_thresh) {
+ return CXL_MBOX_INVALID_INPUT;
+ }
+ alert_config->life_used_warn_thresh = in->life_used_warn_thresh;
+ alert_config->enable_alerts |= CXL_ALERTS_LIFE_USED_WARN_THRESH;
+ }
+
+ if (in->valid_alert_actions & CXL_ALERTS_OVER_TEMP_WARN_THRESH) {
+ /*
+ * CXL r3.2 Table 8-149 The Device Over-Temperature Warning Threshold
+ * shall be less than the the Device Over-Temperature Critical
+ * Alert Threshold.
+ */
+ if (in->over_temp_warn_thresh >=
+ alert_config->over_temp_crit_alert_thresh) {
+ return CXL_MBOX_INVALID_INPUT;
+ }
+ alert_config->over_temp_warn_thresh = in->over_temp_warn_thresh;
+ alert_config->enable_alerts |= CXL_ALERTS_OVER_TEMP_WARN_THRESH;
+ }
+
+ if (in->valid_alert_actions & CXL_ALERTS_UNDER_TEMP_WARN_THRESH) {
+ /*
+ * CXL r3.2 Table 8-149 The Device Under-Temperature Warning Threshold
+ * shall be higher than the the Device Under-Temperature Critical
+ * Alert Threshold.
+ */
+ if (in->under_temp_warn_thresh <=
+ alert_config->under_temp_crit_alert_thresh) {
+ return CXL_MBOX_INVALID_INPUT;
+ }
+ alert_config->under_temp_warn_thresh = in->under_temp_warn_thresh;
+ alert_config->enable_alerts |= CXL_ALERTS_UNDER_TEMP_WARN_THRESH;
+ }
+
+ if (in->valid_alert_actions & CXL_ALERTS_COR_VMEM_ERR_WARN_THRESH) {
+ alert_config->cor_vmem_err_warn_thresh = in->cor_vmem_err_warn_thresh;
+ alert_config->enable_alerts |= CXL_ALERTS_COR_VMEM_ERR_WARN_THRESH;
+ }
+
+ if (in->valid_alert_actions & CXL_ALERTS_COR_PMEM_ERR_WARN_THRESH) {
+ alert_config->cor_pmem_err_warn_thresh = in->cor_pmem_err_warn_thresh;
+ alert_config->enable_alerts |= CXL_ALERTS_COR_PMEM_ERR_WARN_THRESH;
+ }
+ return CXL_MBOX_SUCCESS;
+}
+
/* Perform the actual device zeroing */
static void __do_sanitization(CXLType3Dev *ct3d)
{
@@ -3173,6 +3272,12 @@ static const struct cxl_cmd cxl_cmd_set[256][256] = {
[CCLS][GET_LSA] = { "CCLS_GET_LSA", cmd_ccls_get_lsa, 8, 0 },
[CCLS][SET_LSA] = { "CCLS_SET_LSA", cmd_ccls_set_lsa,
~0, CXL_MBOX_IMMEDIATE_CONFIG_CHANGE | CXL_MBOX_IMMEDIATE_DATA_CHANGE },
+ [HEALTH_INFO_ALERTS][GET_ALERT_CONFIG] = {
+ "HEALTH_INFO_ALERTS_GET_ALERT_CONFIG",
+ cmd_get_alert_config, 0, 0 },
+ [HEALTH_INFO_ALERTS][SET_ALERT_CONFIG] = {
+ "HEALTH_INFO_ALERTS_SET_ALERT_CONFIG",
+ cmd_set_alert_config, 12, CXL_MBOX_IMMEDIATE_POLICY_CHANGE },
[SANITIZE][OVERWRITE] = { "SANITIZE_OVERWRITE", cmd_sanitize_overwrite, 0,
(CXL_MBOX_IMMEDIATE_DATA_CHANGE |
CXL_MBOX_SECURITY_STATE_CHANGE |
diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
index aacd078118..94e7274912 100644
--- a/hw/mem/cxl_type3.c
+++ b/hw/mem/cxl_type3.c
@@ -843,6 +843,19 @@ static DOEProtocol doe_cdat_prot[] = {
{ }
};
+/* Initialize CXL device alerts with default threshold values. */
+static void init_alert_config(CXLType3Dev *ct3d)
+{
+ ct3d->alert_config = (CXLAlertConfig) {
+ .life_used_crit_alert_thresh = 75,
+ .life_used_warn_thresh = 40,
+ .over_temp_crit_alert_thresh = 35,
+ .under_temp_crit_alert_thresh = 10,
+ .over_temp_warn_thresh = 25,
+ .under_temp_warn_thresh = 20
+ };
+}
+
static void ct3_realize(PCIDevice *pci_dev, Error **errp)
{
ERRP_GUARD();
@@ -910,6 +923,7 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
goto err_msix_uninit;
}
+ init_alert_config(ct3d);
pcie_cap_deverr_init(pci_dev);
/* Leave a bit of room for expansion */
rc = pcie_aer_init(pci_dev, PCI_ERR_VER, 0x200, PCI_ERR_SIZEOF, errp);
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 07/27] docs/cxl: Add serial number for persistent-memdev
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (5 preceding siblings ...)
2025-05-14 11:50 ` [PULL 06/27] hw/cxl/cxl-mailbox-utils: CXL CCI Get/Set alert config commands Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 08/27] hw/pci: Do not add ROM BAR for SR-IOV VF Michael S. Tsirkin
` (20 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Yuquan Wang, Jonathan Cameron, Hyeongtak Ji,
Michael Tokarev
From: Yuquan Wang <wangyuquan1236@phytium.com.cn>
Add serial number parameter in the cxl persistent examples.
Signed-off-by: Yuquan Wang <wangyuquan1236@phytium.com.cn>
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Message-Id: <20250305092501.191929-9-Jonathan.Cameron@huawei.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
docs/system/devices/cxl.rst | 18 +++++++++---------
1 file changed, 9 insertions(+), 9 deletions(-)
diff --git a/docs/system/devices/cxl.rst b/docs/system/devices/cxl.rst
index 882b036f5e..e307caf3f8 100644
--- a/docs/system/devices/cxl.rst
+++ b/docs/system/devices/cxl.rst
@@ -308,7 +308,7 @@ A very simple setup with just one directly attached CXL Type 3 Persistent Memory
-object memory-backend-file,id=cxl-lsa1,share=on,mem-path=/tmp/lsa.raw,size=256M \
-device pxb-cxl,bus_nr=12,bus=pcie.0,id=cxl.1 \
-device cxl-rp,port=0,bus=cxl.1,id=root_port13,chassis=0,slot=2 \
- -device cxl-type3,bus=root_port13,persistent-memdev=cxl-mem1,lsa=cxl-lsa1,id=cxl-pmem0 \
+ -device cxl-type3,bus=root_port13,persistent-memdev=cxl-mem1,lsa=cxl-lsa1,id=cxl-pmem0,sn=0x1 \
-M cxl-fmw.0.targets.0=cxl.1,cxl-fmw.0.size=4G
A very simple setup with just one directly attached CXL Type 3 Volatile Memory device::
@@ -349,13 +349,13 @@ the CXL Type3 device directly attached (no switches).::
-device pxb-cxl,bus_nr=12,bus=pcie.0,id=cxl.1 \
-device pxb-cxl,bus_nr=222,bus=pcie.0,id=cxl.2 \
-device cxl-rp,port=0,bus=cxl.1,id=root_port13,chassis=0,slot=2 \
- -device cxl-type3,bus=root_port13,persistent-memdev=cxl-mem1,lsa=cxl-lsa1,id=cxl-pmem0 \
+ -device cxl-type3,bus=root_port13,persistent-memdev=cxl-mem1,lsa=cxl-lsa1,id=cxl-pmem0,sn=0x1 \
-device cxl-rp,port=1,bus=cxl.1,id=root_port14,chassis=0,slot=3 \
- -device cxl-type3,bus=root_port14,persistent-memdev=cxl-mem2,lsa=cxl-lsa2,id=cxl-pmem1 \
+ -device cxl-type3,bus=root_port14,persistent-memdev=cxl-mem2,lsa=cxl-lsa2,id=cxl-pmem1,sn=0x2 \
-device cxl-rp,port=0,bus=cxl.2,id=root_port15,chassis=0,slot=5 \
- -device cxl-type3,bus=root_port15,persistent-memdev=cxl-mem3,lsa=cxl-lsa3,id=cxl-pmem2 \
+ -device cxl-type3,bus=root_port15,persistent-memdev=cxl-mem3,lsa=cxl-lsa3,id=cxl-pmem2,sn=0x3 \
-device cxl-rp,port=1,bus=cxl.2,id=root_port16,chassis=0,slot=6 \
- -device cxl-type3,bus=root_port16,persistent-memdev=cxl-mem4,lsa=cxl-lsa4,id=cxl-pmem3 \
+ -device cxl-type3,bus=root_port16,persistent-memdev=cxl-mem4,lsa=cxl-lsa4,id=cxl-pmem3,sn=0x4 \
-M cxl-fmw.0.targets.0=cxl.1,cxl-fmw.0.targets.1=cxl.2,cxl-fmw.0.size=4G,cxl-fmw.0.interleave-granularity=8k
An example of 4 devices below a switch suitable for 1, 2 or 4 way interleave::
@@ -375,13 +375,13 @@ An example of 4 devices below a switch suitable for 1, 2 or 4 way interleave::
-device cxl-rp,port=1,bus=cxl.1,id=root_port1,chassis=0,slot=1 \
-device cxl-upstream,bus=root_port0,id=us0 \
-device cxl-downstream,port=0,bus=us0,id=swport0,chassis=0,slot=4 \
- -device cxl-type3,bus=swport0,persistent-memdev=cxl-mem0,lsa=cxl-lsa0,id=cxl-pmem0 \
+ -device cxl-type3,bus=swport0,persistent-memdev=cxl-mem0,lsa=cxl-lsa0,id=cxl-pmem0,sn=0x1 \
-device cxl-downstream,port=1,bus=us0,id=swport1,chassis=0,slot=5 \
- -device cxl-type3,bus=swport1,persistent-memdev=cxl-mem1,lsa=cxl-lsa1,id=cxl-pmem1 \
+ -device cxl-type3,bus=swport1,persistent-memdev=cxl-mem1,lsa=cxl-lsa1,id=cxl-pmem1,sn=0x2 \
-device cxl-downstream,port=2,bus=us0,id=swport2,chassis=0,slot=6 \
- -device cxl-type3,bus=swport2,persistent-memdev=cxl-mem2,lsa=cxl-lsa2,id=cxl-pmem2 \
+ -device cxl-type3,bus=swport2,persistent-memdev=cxl-mem2,lsa=cxl-lsa2,id=cxl-pmem2,sn=0x3 \
-device cxl-downstream,port=3,bus=us0,id=swport3,chassis=0,slot=7 \
- -device cxl-type3,bus=swport3,persistent-memdev=cxl-mem3,lsa=cxl-lsa3,id=cxl-pmem3 \
+ -device cxl-type3,bus=swport3,persistent-memdev=cxl-mem3,lsa=cxl-lsa3,id=cxl-pmem3,sn=0x4 \
-M cxl-fmw.0.targets.0=cxl.1,cxl-fmw.0.size=4G,cxl-fmw.0.interleave-granularity=4k
Deprecations
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 08/27] hw/pci: Do not add ROM BAR for SR-IOV VF
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (6 preceding siblings ...)
2025-05-14 11:50 ` [PULL 07/27] docs/cxl: Add serial number for persistent-memdev Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 09/27] hw/pci: Fix SR-IOV VF number calculation Michael S. Tsirkin
` (19 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
A SR-IOV VF cannot have a ROM BAR.
Co-developed-by: Yui Washizu <yui.washidu@gmail.com>
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-1-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/pci/pci.c | 8 ++++++++
1 file changed, 8 insertions(+)
diff --git a/hw/pci/pci.c b/hw/pci/pci.c
index fe38c4c028..6d9d3ce90f 100644
--- a/hw/pci/pci.c
+++ b/hw/pci/pci.c
@@ -2522,6 +2522,14 @@ static void pci_add_option_rom(PCIDevice *pdev, bool is_default_rom,
return;
}
+ if (pci_is_vf(pdev)) {
+ if (pdev->rom_bar > 0) {
+ error_setg(errp, "ROM BAR cannot be enabled for SR-IOV VF");
+ }
+
+ return;
+ }
+
if (load_file || pdev->romsize == UINT32_MAX) {
path = qemu_find_file(QEMU_FILE_TYPE_BIOS, pdev->romfile);
if (path == NULL) {
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 09/27] hw/pci: Fix SR-IOV VF number calculation
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (7 preceding siblings ...)
2025-05-14 11:50 ` [PULL 08/27] hw/pci: Do not add ROM BAR for SR-IOV VF Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 10/27] pcie_sriov: Ensure PF and VF are mutually exclusive Michael S. Tsirkin
` (18 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
pci_config_get_bar_addr() had a division by vf_stride. vf_stride needs
to be non-zero when there are multiple VFs, but the specification does
not prohibit to make it zero when there is only one VF.
Do not perform the division for the first VF to avoid division by zero.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-2-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/pci/pci.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/hw/pci/pci.c b/hw/pci/pci.c
index 6d9d3ce90f..039cf1e256 100644
--- a/hw/pci/pci.c
+++ b/hw/pci/pci.c
@@ -1594,7 +1594,11 @@ static pcibus_t pci_config_get_bar_addr(PCIDevice *d, int reg,
pci_get_word(pf->config + sriov_cap + PCI_SRIOV_VF_OFFSET);
uint16_t vf_stride =
pci_get_word(pf->config + sriov_cap + PCI_SRIOV_VF_STRIDE);
- uint32_t vf_num = (d->devfn - (pf->devfn + vf_offset)) / vf_stride;
+ uint32_t vf_num = d->devfn - (pf->devfn + vf_offset);
+
+ if (vf_num) {
+ vf_num /= vf_stride;
+ }
if (type & PCI_BASE_ADDRESS_MEM_TYPE_64) {
new_addr = pci_get_quad(pf->config + bar);
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 10/27] pcie_sriov: Ensure PF and VF are mutually exclusive
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (8 preceding siblings ...)
2025-05-14 11:50 ` [PULL 09/27] hw/pci: Fix SR-IOV VF number calculation Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 11/27] pcie_sriov: Check PCI Express for SR-IOV PF Michael S. Tsirkin
` (17 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
A device cannot be a SR-IOV PF and a VF at the same time.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-3-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/pci/pcie_sriov.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/hw/pci/pcie_sriov.c b/hw/pci/pcie_sriov.c
index 1eb4358256..109b2ebccc 100644
--- a/hw/pci/pcie_sriov.c
+++ b/hw/pci/pcie_sriov.c
@@ -42,6 +42,11 @@ bool pcie_sriov_pf_init(PCIDevice *dev, uint16_t offset,
uint8_t *cfg = dev->config + offset;
uint8_t *wmask;
+ if (pci_is_vf(dev)) {
+ error_setg(errp, "a device cannot be a SR-IOV PF and a VF at the same time");
+ return false;
+ }
+
if (total_vfs &&
(uint32_t)devfn + (uint32_t)(total_vfs - 1) * vf_stride >= PCI_DEVFN_MAX) {
error_setg(errp, "VF addr overflows");
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 11/27] pcie_sriov: Check PCI Express for SR-IOV PF
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (9 preceding siblings ...)
2025-05-14 11:50 ` [PULL 10/27] pcie_sriov: Ensure PF and VF are mutually exclusive Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 12/27] pcie_sriov: Allow user to create SR-IOV device Michael S. Tsirkin
` (16 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
SR-IOV requires PCI Express.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-4-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/pci/pcie_sriov.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/hw/pci/pcie_sriov.c b/hw/pci/pcie_sriov.c
index 109b2ebccc..a5b546abe8 100644
--- a/hw/pci/pcie_sriov.c
+++ b/hw/pci/pcie_sriov.c
@@ -42,6 +42,11 @@ bool pcie_sriov_pf_init(PCIDevice *dev, uint16_t offset,
uint8_t *cfg = dev->config + offset;
uint8_t *wmask;
+ if (!pci_is_express(dev)) {
+ error_setg(errp, "PCI Express is required for SR-IOV PF");
+ return false;
+ }
+
if (pci_is_vf(dev)) {
error_setg(errp, "a device cannot be a SR-IOV PF and a VF at the same time");
return false;
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 12/27] pcie_sriov: Allow user to create SR-IOV device
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (10 preceding siblings ...)
2025-05-14 11:50 ` [PULL 11/27] pcie_sriov: Check PCI Express for SR-IOV PF Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 13/27] virtio-pci: Implement SR-IOV PF Michael S. Tsirkin
` (15 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
A user can create a SR-IOV device by specifying the PF with the
sriov-pf property of the VFs. The VFs must be added before the PF.
A user-creatable VF must have PCIDeviceClass::sriov_vf_user_creatable
set. Such a VF cannot refer to the PF because it is created before the
PF.
A PF that user-creatable VFs can be attached calls
pcie_sriov_pf_init_from_user_created_vfs() during realization and
pcie_sriov_pf_exit() when exiting.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-5-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/pci/pci_device.h | 6 +-
include/hw/pci/pcie_sriov.h | 18 +++
hw/pci/pci.c | 62 +++++---
hw/pci/pcie_sriov.c | 288 ++++++++++++++++++++++++++++--------
4 files changed, 291 insertions(+), 83 deletions(-)
diff --git a/include/hw/pci/pci_device.h b/include/hw/pci/pci_device.h
index 345b12eaac..e41d95b0b0 100644
--- a/include/hw/pci/pci_device.h
+++ b/include/hw/pci/pci_device.h
@@ -38,6 +38,8 @@ struct PCIDeviceClass {
uint16_t subsystem_id; /* only for header type = 0 */
const char *romfile; /* rom bar */
+
+ bool sriov_vf_user_creatable;
};
enum PCIReqIDType {
@@ -177,6 +179,8 @@ struct PCIDevice {
* realizing the device.
*/
uint32_t max_bounce_buffer_size;
+
+ char *sriov_pf;
};
static inline int pci_intx(PCIDevice *pci_dev)
@@ -209,7 +213,7 @@ static inline int pci_is_express_downstream_port(const PCIDevice *d)
static inline int pci_is_vf(const PCIDevice *d)
{
- return d->exp.sriov_vf.pf != NULL;
+ return d->sriov_pf || d->exp.sriov_vf.pf != NULL;
}
static inline uint32_t pci_config_size(const PCIDevice *d)
diff --git a/include/hw/pci/pcie_sriov.h b/include/hw/pci/pcie_sriov.h
index c5d2d318d3..f75b8f22ee 100644
--- a/include/hw/pci/pcie_sriov.h
+++ b/include/hw/pci/pcie_sriov.h
@@ -18,6 +18,7 @@
typedef struct PCIESriovPF {
uint8_t vf_bar_type[PCI_NUM_REGIONS]; /* Store type for each VF bar */
PCIDevice **vf; /* Pointer to an array of num_vfs VF devices */
+ bool vf_user_created; /* If VFs are created by user */
} PCIESriovPF;
typedef struct PCIESriovVF {
@@ -40,6 +41,23 @@ void pcie_sriov_pf_init_vf_bar(PCIDevice *dev, int region_num,
void pcie_sriov_vf_register_bar(PCIDevice *dev, int region_num,
MemoryRegion *memory);
+/**
+ * pcie_sriov_pf_init_from_user_created_vfs() - Initialize PF with user-created
+ * VFs.
+ * @dev: A PCIe device being realized.
+ * @offset: The offset of the SR-IOV capability.
+ * @errp: pointer to Error*, to store an error if it happens.
+ *
+ * Return: The size of added capability. 0 if the user did not create VFs.
+ * -1 if failed.
+ */
+int16_t pcie_sriov_pf_init_from_user_created_vfs(PCIDevice *dev,
+ uint16_t offset,
+ Error **errp);
+
+bool pcie_sriov_register_device(PCIDevice *dev, Error **errp);
+void pcie_sriov_unregister_device(PCIDevice *dev);
+
/*
* Default (minimal) page size support values
* as required by the SR/IOV standard:
diff --git a/hw/pci/pci.c b/hw/pci/pci.c
index 039cf1e256..de07096347 100644
--- a/hw/pci/pci.c
+++ b/hw/pci/pci.c
@@ -101,6 +101,7 @@ static const Property pci_props[] = {
QEMU_PCIE_ARI_NEXTFN_1_BITNR, false),
DEFINE_PROP_SIZE32("x-max-bounce-buffer-size", PCIDevice,
max_bounce_buffer_size, DEFAULT_MAX_BOUNCE_BUFFER_SIZE),
+ DEFINE_PROP_STRING("sriov-pf", PCIDevice, sriov_pf),
DEFINE_PROP_BIT("x-pcie-ext-tag", PCIDevice, cap_present,
QEMU_PCIE_EXT_TAG_BITNR, true),
{ .name = "busnr", .info = &prop_pci_busnr },
@@ -1112,13 +1113,8 @@ static void pci_init_multifunction(PCIBus *bus, PCIDevice *dev, Error **errp)
dev->config[PCI_HEADER_TYPE] |= PCI_HEADER_TYPE_MULTI_FUNCTION;
}
- /*
- * With SR/IOV and ARI, a device at function 0 need not be a multifunction
- * device, as it may just be a VF that ended up with function 0 in
- * the legacy PCI interpretation. Avoid failing in such cases:
- */
- if (pci_is_vf(dev) &&
- dev->exp.sriov_vf.pf->cap_present & QEMU_PCI_CAP_MULTIFUNCTION) {
+ /* SR/IOV is not handled here. */
+ if (pci_is_vf(dev)) {
return;
}
@@ -1151,7 +1147,8 @@ static void pci_init_multifunction(PCIBus *bus, PCIDevice *dev, Error **errp)
}
/* function 0 indicates single function, so function > 0 must be NULL */
for (func = 1; func < PCI_FUNC_MAX; ++func) {
- if (bus->devices[PCI_DEVFN(slot, func)]) {
+ PCIDevice *device = bus->devices[PCI_DEVFN(slot, func)];
+ if (device && !pci_is_vf(device)) {
error_setg(errp, "PCI: %x.0 indicates single function, "
"but %x.%x is already populated.",
slot, slot, func);
@@ -1439,6 +1436,7 @@ static void pci_qdev_unrealize(DeviceState *dev)
pci_unregister_io_regions(pci_dev);
pci_del_option_rom(pci_dev);
+ pcie_sriov_unregister_device(pci_dev);
if (pc->exit) {
pc->exit(pci_dev);
@@ -1470,7 +1468,6 @@ void pci_register_bar(PCIDevice *pci_dev, int region_num,
pcibus_t size = memory_region_size(memory);
uint8_t hdr_type;
- assert(!pci_is_vf(pci_dev)); /* VFs must use pcie_sriov_vf_register_bar */
assert(region_num >= 0);
assert(region_num < PCI_NUM_REGIONS);
assert(is_power_of_2(size));
@@ -1482,7 +1479,6 @@ void pci_register_bar(PCIDevice *pci_dev, int region_num,
r = &pci_dev->io_regions[region_num];
assert(!r->size);
- r->addr = PCI_BAR_UNMAPPED;
r->size = size;
r->type = type;
r->memory = memory;
@@ -1490,22 +1486,35 @@ void pci_register_bar(PCIDevice *pci_dev, int region_num,
? pci_get_bus(pci_dev)->address_space_io
: pci_get_bus(pci_dev)->address_space_mem;
- wmask = ~(size - 1);
- if (region_num == PCI_ROM_SLOT) {
- /* ROM enable bit is writable */
- wmask |= PCI_ROM_ADDRESS_ENABLE;
- }
+ if (pci_is_vf(pci_dev)) {
+ PCIDevice *pf = pci_dev->exp.sriov_vf.pf;
+ assert(!pf || type == pf->exp.sriov_pf.vf_bar_type[region_num]);
- addr = pci_bar(pci_dev, region_num);
- pci_set_long(pci_dev->config + addr, type);
-
- if (!(r->type & PCI_BASE_ADDRESS_SPACE_IO) &&
- r->type & PCI_BASE_ADDRESS_MEM_TYPE_64) {
- pci_set_quad(pci_dev->wmask + addr, wmask);
- pci_set_quad(pci_dev->cmask + addr, ~0ULL);
+ r->addr = pci_bar_address(pci_dev, region_num, r->type, r->size);
+ if (r->addr != PCI_BAR_UNMAPPED) {
+ memory_region_add_subregion_overlap(r->address_space,
+ r->addr, r->memory, 1);
+ }
} else {
- pci_set_long(pci_dev->wmask + addr, wmask & 0xffffffff);
- pci_set_long(pci_dev->cmask + addr, 0xffffffff);
+ r->addr = PCI_BAR_UNMAPPED;
+
+ wmask = ~(size - 1);
+ if (region_num == PCI_ROM_SLOT) {
+ /* ROM enable bit is writable */
+ wmask |= PCI_ROM_ADDRESS_ENABLE;
+ }
+
+ addr = pci_bar(pci_dev, region_num);
+ pci_set_long(pci_dev->config + addr, type);
+
+ if (!(r->type & PCI_BASE_ADDRESS_SPACE_IO) &&
+ r->type & PCI_BASE_ADDRESS_MEM_TYPE_64) {
+ pci_set_quad(pci_dev->wmask + addr, wmask);
+ pci_set_quad(pci_dev->cmask + addr, ~0ULL);
+ } else {
+ pci_set_long(pci_dev->wmask + addr, wmask & 0xffffffff);
+ pci_set_long(pci_dev->cmask + addr, 0xffffffff);
+ }
}
}
@@ -2272,6 +2281,11 @@ static void pci_qdev_realize(DeviceState *qdev, Error **errp)
}
}
+ if (!pcie_sriov_register_device(pci_dev, errp)) {
+ pci_qdev_unrealize(DEVICE(pci_dev));
+ return;
+ }
+
/*
* A PCIe Downstream Port that do not have ARI Forwarding enabled must
* associate only Device 0 with the device attached to the bus
diff --git a/hw/pci/pcie_sriov.c b/hw/pci/pcie_sriov.c
index a5b546abe8..08f707e847 100644
--- a/hw/pci/pcie_sriov.c
+++ b/hw/pci/pcie_sriov.c
@@ -15,11 +15,12 @@
#include "hw/pci/pcie.h"
#include "hw/pci/pci_bus.h"
#include "hw/qdev-properties.h"
-#include "qemu/error-report.h"
#include "qemu/range.h"
#include "qapi/error.h"
#include "trace.h"
+static GHashTable *pfs;
+
static void unparent_vfs(PCIDevice *dev, uint16_t total_vfs)
{
for (uint16_t i = 0; i < total_vfs; i++) {
@@ -31,13 +32,43 @@ static void unparent_vfs(PCIDevice *dev, uint16_t total_vfs)
dev->exp.sriov_pf.vf = NULL;
}
-bool pcie_sriov_pf_init(PCIDevice *dev, uint16_t offset,
- const char *vfname, uint16_t vf_dev_id,
- uint16_t init_vfs, uint16_t total_vfs,
- uint16_t vf_offset, uint16_t vf_stride,
- Error **errp)
+static void register_vfs(PCIDevice *dev)
+{
+ uint16_t num_vfs;
+ uint16_t i;
+ uint16_t sriov_cap = dev->exp.sriov_cap;
+
+ assert(sriov_cap > 0);
+ num_vfs = pci_get_word(dev->config + sriov_cap + PCI_SRIOV_NUM_VF);
+
+ trace_sriov_register_vfs(dev->name, PCI_SLOT(dev->devfn),
+ PCI_FUNC(dev->devfn), num_vfs);
+ for (i = 0; i < num_vfs; i++) {
+ pci_set_enabled(dev->exp.sriov_pf.vf[i], true);
+ }
+
+ pci_set_word(dev->wmask + sriov_cap + PCI_SRIOV_NUM_VF, 0);
+}
+
+static void unregister_vfs(PCIDevice *dev)
+{
+ uint8_t *cfg = dev->config + dev->exp.sriov_cap;
+ uint16_t i;
+
+ trace_sriov_unregister_vfs(dev->name, PCI_SLOT(dev->devfn),
+ PCI_FUNC(dev->devfn));
+ for (i = 0; i < pci_get_word(cfg + PCI_SRIOV_TOTAL_VF); i++) {
+ pci_set_enabled(dev->exp.sriov_pf.vf[i], false);
+ }
+
+ pci_set_word(dev->wmask + dev->exp.sriov_cap + PCI_SRIOV_NUM_VF, 0xffff);
+}
+
+static bool pcie_sriov_pf_init_common(PCIDevice *dev, uint16_t offset,
+ uint16_t vf_dev_id, uint16_t init_vfs,
+ uint16_t total_vfs, uint16_t vf_offset,
+ uint16_t vf_stride, Error **errp)
{
- BusState *bus = qdev_get_parent_bus(&dev->qdev);
int32_t devfn = dev->devfn + vf_offset;
uint8_t *cfg = dev->config + offset;
uint8_t *wmask;
@@ -94,6 +125,28 @@ bool pcie_sriov_pf_init(PCIDevice *dev, uint16_t offset,
qdev_prop_set_bit(&dev->qdev, "multifunction", true);
+ return true;
+}
+
+bool pcie_sriov_pf_init(PCIDevice *dev, uint16_t offset,
+ const char *vfname, uint16_t vf_dev_id,
+ uint16_t init_vfs, uint16_t total_vfs,
+ uint16_t vf_offset, uint16_t vf_stride,
+ Error **errp)
+{
+ BusState *bus = qdev_get_parent_bus(&dev->qdev);
+ int32_t devfn = dev->devfn + vf_offset;
+
+ if (pfs && g_hash_table_contains(pfs, dev->qdev.id)) {
+ error_setg(errp, "attaching user-created SR-IOV VF unsupported");
+ return false;
+ }
+
+ if (!pcie_sriov_pf_init_common(dev, offset, vf_dev_id, init_vfs,
+ total_vfs, vf_offset, vf_stride, errp)) {
+ return false;
+ }
+
dev->exp.sriov_pf.vf = g_new(PCIDevice *, total_vfs);
for (uint16_t i = 0; i < total_vfs; i++) {
@@ -123,7 +176,22 @@ void pcie_sriov_pf_exit(PCIDevice *dev)
{
uint8_t *cfg = dev->config + dev->exp.sriov_cap;
- unparent_vfs(dev, pci_get_word(cfg + PCI_SRIOV_TOTAL_VF));
+ if (dev->exp.sriov_pf.vf_user_created) {
+ uint16_t ven_id = pci_get_word(dev->config + PCI_VENDOR_ID);
+ uint16_t total_vfs = pci_get_word(dev->config + PCI_SRIOV_TOTAL_VF);
+ uint16_t vf_dev_id = pci_get_word(dev->config + PCI_SRIOV_VF_DID);
+
+ unregister_vfs(dev);
+
+ for (uint16_t i = 0; i < total_vfs; i++) {
+ dev->exp.sriov_pf.vf[i]->exp.sriov_vf.pf = NULL;
+
+ pci_config_set_vendor_id(dev->exp.sriov_pf.vf[i]->config, ven_id);
+ pci_config_set_device_id(dev->exp.sriov_pf.vf[i]->config, vf_dev_id);
+ }
+ } else {
+ unparent_vfs(dev, pci_get_word(cfg + PCI_SRIOV_TOTAL_VF));
+ }
}
void pcie_sriov_pf_init_vf_bar(PCIDevice *dev, int region_num,
@@ -156,69 +224,173 @@ void pcie_sriov_pf_init_vf_bar(PCIDevice *dev, int region_num,
void pcie_sriov_vf_register_bar(PCIDevice *dev, int region_num,
MemoryRegion *memory)
{
- PCIIORegion *r;
- PCIBus *bus = pci_get_bus(dev);
uint8_t type;
- pcibus_t size = memory_region_size(memory);
- assert(pci_is_vf(dev)); /* PFs must use pci_register_bar */
- assert(region_num >= 0);
- assert(region_num < PCI_NUM_REGIONS);
+ assert(dev->exp.sriov_vf.pf);
type = dev->exp.sriov_vf.pf->exp.sriov_pf.vf_bar_type[region_num];
- if (!is_power_of_2(size)) {
- error_report("%s: PCI region size must be a power"
- " of two - type=0x%x, size=0x%"FMT_PCIBUS,
- __func__, type, size);
- exit(1);
- }
-
- r = &dev->io_regions[region_num];
- r->memory = memory;
- r->address_space =
- type & PCI_BASE_ADDRESS_SPACE_IO
- ? bus->address_space_io
- : bus->address_space_mem;
- r->size = size;
- r->type = type;
-
- r->addr = pci_bar_address(dev, region_num, r->type, r->size);
- if (r->addr != PCI_BAR_UNMAPPED) {
- memory_region_add_subregion_overlap(r->address_space,
- r->addr, r->memory, 1);
- }
+ return pci_register_bar(dev, region_num, type, memory);
}
-static void register_vfs(PCIDevice *dev)
+static gint compare_vf_devfns(gconstpointer a, gconstpointer b)
{
- uint16_t num_vfs;
- uint16_t i;
- uint16_t sriov_cap = dev->exp.sriov_cap;
-
- assert(sriov_cap > 0);
- num_vfs = pci_get_word(dev->config + sriov_cap + PCI_SRIOV_NUM_VF);
-
- trace_sriov_register_vfs(dev->name, PCI_SLOT(dev->devfn),
- PCI_FUNC(dev->devfn), num_vfs);
- for (i = 0; i < num_vfs; i++) {
- pci_set_enabled(dev->exp.sriov_pf.vf[i], true);
- }
-
- pci_set_word(dev->wmask + sriov_cap + PCI_SRIOV_NUM_VF, 0);
+ return (*(PCIDevice **)a)->devfn - (*(PCIDevice **)b)->devfn;
}
-static void unregister_vfs(PCIDevice *dev)
+int16_t pcie_sriov_pf_init_from_user_created_vfs(PCIDevice *dev,
+ uint16_t offset,
+ Error **errp)
{
- uint8_t *cfg = dev->config + dev->exp.sriov_cap;
+ GPtrArray *pf;
+ PCIDevice **vfs;
+ BusState *bus = qdev_get_parent_bus(DEVICE(dev));
+ uint16_t ven_id = pci_get_word(dev->config + PCI_VENDOR_ID);
+ uint16_t vf_dev_id;
+ uint16_t vf_offset;
+ uint16_t vf_stride;
uint16_t i;
- trace_sriov_unregister_vfs(dev->name, PCI_SLOT(dev->devfn),
- PCI_FUNC(dev->devfn));
- for (i = 0; i < pci_get_word(cfg + PCI_SRIOV_TOTAL_VF); i++) {
- pci_set_enabled(dev->exp.sriov_pf.vf[i], false);
+ if (!pfs || !dev->qdev.id) {
+ return 0;
}
- pci_set_word(dev->wmask + dev->exp.sriov_cap + PCI_SRIOV_NUM_VF, 0xffff);
+ pf = g_hash_table_lookup(pfs, dev->qdev.id);
+ if (!pf) {
+ return 0;
+ }
+
+ if (pf->len > UINT16_MAX) {
+ error_setg(errp, "too many VFs");
+ return -1;
+ }
+
+ g_ptr_array_sort(pf, compare_vf_devfns);
+ vfs = (void *)pf->pdata;
+
+ if (vfs[0]->devfn <= dev->devfn) {
+ error_setg(errp, "a VF function number is less than the PF function number");
+ return -1;
+ }
+
+ vf_dev_id = pci_get_word(vfs[0]->config + PCI_DEVICE_ID);
+ vf_offset = vfs[0]->devfn - dev->devfn;
+ vf_stride = pf->len < 2 ? 0 : vfs[1]->devfn - vfs[0]->devfn;
+
+ for (i = 0; i < pf->len; i++) {
+ if (bus != qdev_get_parent_bus(&vfs[i]->qdev)) {
+ error_setg(errp, "SR-IOV VF parent bus mismatches with PF");
+ return -1;
+ }
+
+ if (ven_id != pci_get_word(vfs[i]->config + PCI_VENDOR_ID)) {
+ error_setg(errp, "SR-IOV VF vendor ID mismatches with PF");
+ return -1;
+ }
+
+ if (vf_dev_id != pci_get_word(vfs[i]->config + PCI_DEVICE_ID)) {
+ error_setg(errp, "inconsistent SR-IOV VF device IDs");
+ return -1;
+ }
+
+ for (size_t j = 0; j < PCI_NUM_REGIONS; j++) {
+ if (vfs[i]->io_regions[j].size != vfs[0]->io_regions[j].size ||
+ vfs[i]->io_regions[j].type != vfs[0]->io_regions[j].type) {
+ error_setg(errp, "inconsistent SR-IOV BARs");
+ return -1;
+ }
+ }
+
+ if (vfs[i]->devfn - vfs[0]->devfn != vf_stride * i) {
+ error_setg(errp, "inconsistent SR-IOV stride");
+ return -1;
+ }
+ }
+
+ if (!pcie_sriov_pf_init_common(dev, offset, vf_dev_id, pf->len,
+ pf->len, vf_offset, vf_stride, errp)) {
+ return -1;
+ }
+
+ for (i = 0; i < pf->len; i++) {
+ vfs[i]->exp.sriov_vf.pf = dev;
+ vfs[i]->exp.sriov_vf.vf_number = i;
+
+ /* set vid/did according to sr/iov spec - they are not used */
+ pci_config_set_vendor_id(vfs[i]->config, 0xffff);
+ pci_config_set_device_id(vfs[i]->config, 0xffff);
+ }
+
+ dev->exp.sriov_pf.vf = vfs;
+ dev->exp.sriov_pf.vf_user_created = true;
+
+ for (i = 0; i < PCI_NUM_REGIONS; i++) {
+ PCIIORegion *region = &vfs[0]->io_regions[i];
+
+ if (region->size) {
+ pcie_sriov_pf_init_vf_bar(dev, i, region->type, region->size);
+ }
+ }
+
+ return PCI_EXT_CAP_SRIOV_SIZEOF;
+}
+
+bool pcie_sriov_register_device(PCIDevice *dev, Error **errp)
+{
+ if (!dev->exp.sriov_pf.vf && dev->qdev.id &&
+ pfs && g_hash_table_contains(pfs, dev->qdev.id)) {
+ error_setg(errp, "attaching user-created SR-IOV VF unsupported");
+ return false;
+ }
+
+ if (dev->sriov_pf) {
+ PCIDevice *pci_pf;
+ GPtrArray *pf;
+
+ if (!PCI_DEVICE_GET_CLASS(dev)->sriov_vf_user_creatable) {
+ error_setg(errp, "user cannot create SR-IOV VF with this device type");
+ return false;
+ }
+
+ if (!pci_is_express(dev)) {
+ error_setg(errp, "PCI Express is required for SR-IOV VF");
+ return false;
+ }
+
+ if (!pci_qdev_find_device(dev->sriov_pf, &pci_pf)) {
+ error_setg(errp, "PCI device specified as SR-IOV PF already exists");
+ return false;
+ }
+
+ if (!pfs) {
+ pfs = g_hash_table_new_full(g_str_hash, g_str_equal, g_free, NULL);
+ }
+
+ pf = g_hash_table_lookup(pfs, dev->sriov_pf);
+ if (!pf) {
+ pf = g_ptr_array_new();
+ g_hash_table_insert(pfs, g_strdup(dev->sriov_pf), pf);
+ }
+
+ g_ptr_array_add(pf, dev);
+ }
+
+ return true;
+}
+
+void pcie_sriov_unregister_device(PCIDevice *dev)
+{
+ if (dev->sriov_pf && pfs) {
+ GPtrArray *pf = g_hash_table_lookup(pfs, dev->sriov_pf);
+
+ if (pf) {
+ g_ptr_array_remove_fast(pf, dev);
+
+ if (!pf->len) {
+ g_hash_table_remove(pfs, dev->sriov_pf);
+ g_ptr_array_free(pf, FALSE);
+ }
+ }
+ }
}
void pcie_sriov_config_write(PCIDevice *dev, uint32_t address,
@@ -314,7 +486,7 @@ void pcie_sriov_pf_add_sup_pgsize(PCIDevice *dev, uint16_t opt_sup_pgsize)
uint16_t pcie_sriov_vf_number(PCIDevice *dev)
{
- assert(pci_is_vf(dev));
+ assert(dev->exp.sriov_vf.pf);
return dev->exp.sriov_vf.vf_number;
}
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 13/27] virtio-pci: Implement SR-IOV PF
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (11 preceding siblings ...)
2025-05-14 11:50 ` [PULL 12/27] pcie_sriov: Allow user to create SR-IOV device Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 14/27] virtio-net: Implement SR-IOV VF Michael S. Tsirkin
` (14 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin
From: Akihiko Odaki <akihiko.odaki@daynix.com>
Allow user to attach SR-IOV VF to a virtio-pci PF.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-6-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/virtio/virtio-pci.h | 1 +
hw/virtio/virtio-pci.c | 20 +++++++++++++++-----
2 files changed, 16 insertions(+), 5 deletions(-)
diff --git a/include/hw/virtio/virtio-pci.h b/include/hw/virtio/virtio-pci.h
index 31ec144509..1dbc3851b0 100644
--- a/include/hw/virtio/virtio-pci.h
+++ b/include/hw/virtio/virtio-pci.h
@@ -155,6 +155,7 @@ struct VirtIOPCIProxy {
uint32_t modern_io_bar_idx;
uint32_t modern_mem_bar_idx;
int config_cap;
+ uint16_t last_pcie_cap_offset;
uint32_t flags;
bool disable_modern;
bool ignore_backend_features;
diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
index 0fa8fe4955..fee65d3645 100644
--- a/hw/virtio/virtio-pci.c
+++ b/hw/virtio/virtio-pci.c
@@ -1962,6 +1962,7 @@ static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
uint8_t *config;
uint32_t size;
VirtIODevice *vdev = virtio_bus_get_device(bus);
+ int16_t res;
/*
* Virtio capabilities present without
@@ -2109,6 +2110,14 @@ static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
pci_register_bar(&proxy->pci_dev, proxy->legacy_io_bar_idx,
PCI_BASE_ADDRESS_SPACE_IO, &proxy->bar);
}
+
+ res = pcie_sriov_pf_init_from_user_created_vfs(&proxy->pci_dev,
+ proxy->last_pcie_cap_offset,
+ errp);
+ if (res > 0) {
+ proxy->last_pcie_cap_offset += res;
+ virtio_add_feature(&vdev->host_features, VIRTIO_F_SR_IOV);
+ }
}
static void virtio_pci_device_unplugged(DeviceState *d)
@@ -2199,7 +2208,7 @@ static void virtio_pci_realize(PCIDevice *pci_dev, Error **errp)
if (pcie_port && pci_is_express(pci_dev)) {
int pos;
- uint16_t last_pcie_cap_offset = PCI_CONFIG_SPACE_SIZE;
+ proxy->last_pcie_cap_offset = PCI_CONFIG_SPACE_SIZE;
pos = pcie_endpoint_cap_init(pci_dev, 0);
assert(pos > 0);
@@ -2216,9 +2225,9 @@ static void virtio_pci_realize(PCIDevice *pci_dev, Error **errp)
pci_set_word(pci_dev->config + pos + PCI_PM_PMC, 0x3);
if (proxy->flags & VIRTIO_PCI_FLAG_AER) {
- pcie_aer_init(pci_dev, PCI_ERR_VER, last_pcie_cap_offset,
+ pcie_aer_init(pci_dev, PCI_ERR_VER, proxy->last_pcie_cap_offset,
PCI_ERR_SIZEOF, NULL);
- last_pcie_cap_offset += PCI_ERR_SIZEOF;
+ proxy->last_pcie_cap_offset += PCI_ERR_SIZEOF;
}
if (proxy->flags & VIRTIO_PCI_FLAG_INIT_DEVERR) {
@@ -2243,9 +2252,9 @@ static void virtio_pci_realize(PCIDevice *pci_dev, Error **errp)
}
if (proxy->flags & VIRTIO_PCI_FLAG_ATS) {
- pcie_ats_init(pci_dev, last_pcie_cap_offset,
+ pcie_ats_init(pci_dev, proxy->last_pcie_cap_offset,
proxy->flags & VIRTIO_PCI_FLAG_ATS_PAGE_ALIGNED);
- last_pcie_cap_offset += PCI_EXT_CAP_ATS_SIZEOF;
+ proxy->last_pcie_cap_offset += PCI_EXT_CAP_ATS_SIZEOF;
}
if (proxy->flags & VIRTIO_PCI_FLAG_INIT_FLR) {
@@ -2273,6 +2282,7 @@ static void virtio_pci_exit(PCIDevice *pci_dev)
!pci_bus_is_root(pci_get_bus(pci_dev));
bool modern_pio = proxy->flags & VIRTIO_PCI_FLAG_MODERN_PIO_NOTIFY;
+ pcie_sriov_pf_exit(&proxy->pci_dev);
msix_uninit_exclusive_bar(pci_dev);
if (proxy->flags & VIRTIO_PCI_FLAG_AER && pcie_port &&
pci_is_express(pci_dev)) {
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 14/27] virtio-net: Implement SR-IOV VF
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (12 preceding siblings ...)
2025-05-14 11:50 ` [PULL 13/27] virtio-pci: Implement SR-IOV PF Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 15/27] docs: Document composable SR-IOV device Michael S. Tsirkin
` (13 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin
From: Akihiko Odaki <akihiko.odaki@daynix.com>
A virtio-net device can be added as a SR-IOV VF to another virtio-pci
device that will be the PF.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-7-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/virtio/virtio-net-pci.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/hw/virtio/virtio-net-pci.c b/hw/virtio/virtio-net-pci.c
index 8cf9788bc3..f857a84f11 100644
--- a/hw/virtio/virtio-net-pci.c
+++ b/hw/virtio/virtio-net-pci.c
@@ -74,6 +74,7 @@ static void virtio_net_pci_class_init(ObjectClass *klass, const void *data)
k->device_id = PCI_DEVICE_ID_VIRTIO_NET;
k->revision = VIRTIO_PCI_ABI_VERSION;
k->class_id = PCI_CLASS_NETWORK_ETHERNET;
+ k->sriov_vf_user_creatable = true;
set_bit(DEVICE_CATEGORY_NETWORK, dc->categories);
device_class_set_props(dc, virtio_net_properties);
vpciklass->realize = virtio_net_pci_realize;
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 15/27] docs: Document composable SR-IOV device
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (13 preceding siblings ...)
2025-05-14 11:50 ` [PULL 14/27] virtio-net: Implement SR-IOV VF Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 16/27] pcie_sriov: Make a PCI device with user-created VF ARI-capable Michael S. Tsirkin
` (12 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-8-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
MAINTAINERS | 1 +
docs/system/index.rst | 1 +
docs/system/sriov.rst | 36 ++++++++++++++++++++++++++++++++++++
3 files changed, 38 insertions(+)
create mode 100644 docs/system/sriov.rst
diff --git a/MAINTAINERS b/MAINTAINERS
index 6dacd6d004..b579358cdd 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -2058,6 +2058,7 @@ F: hw/pci-bridge/*
F: qapi/pci.json
F: docs/pci*
F: docs/specs/*pci*
+F: docs/system/sriov.rst
PCIE DOE
M: Huai-Cheng Kuo <hchkuo@avery-design.com.tw>
diff --git a/docs/system/index.rst b/docs/system/index.rst
index c21065e519..718e9d3c56 100644
--- a/docs/system/index.rst
+++ b/docs/system/index.rst
@@ -39,3 +39,4 @@ or Hypervisor.Framework.
multi-process
confidential-guest-support
vm-templating
+ sriov
diff --git a/docs/system/sriov.rst b/docs/system/sriov.rst
new file mode 100644
index 0000000000..a851a66a4b
--- /dev/null
+++ b/docs/system/sriov.rst
@@ -0,0 +1,36 @@
+.. SPDX-License-Identifier: GPL-2.0-or-later
+
+Compsable SR-IOV device
+=======================
+
+SR-IOV (Single Root I/O Virtualization) is an optional extended capability of a
+PCI Express device. It allows a single physical function (PF) to appear as
+multiple virtual functions (VFs) for the main purpose of eliminating software
+overhead in I/O from virtual machines.
+
+There are devices with predefined SR-IOV configurations, but it is also possible
+to compose an SR-IOV device yourself. Composing an SR-IOV device is currently
+only supported by virtio-net-pci.
+
+Users can configure an SR-IOV-capable virtio-net device by adding
+virtio-net-pci functions to a bus. Below is a command line example:
+
+.. code-block:: shell
+
+ -netdev user,id=n -netdev user,id=o
+ -netdev user,id=p -netdev user,id=q
+ -device pcie-root-port,id=b
+ -device virtio-net-pci,bus=b,addr=0x0.0x3,netdev=q,sriov-pf=f
+ -device virtio-net-pci,bus=b,addr=0x0.0x2,netdev=p,sriov-pf=f
+ -device virtio-net-pci,bus=b,addr=0x0.0x1,netdev=o,sriov-pf=f
+ -device virtio-net-pci,bus=b,addr=0x0.0x0,netdev=n,id=f
+
+The VFs specify the paired PF with ``sriov-pf`` property. The PF must be
+added after all VFs. It is the user's responsibility to ensure that VFs have
+function numbers larger than one of the PF, and that the function numbers
+have a consistent stride.
+
+You may also need to perform additional steps to activate the SR-IOV feature on
+your guest. For Linux, refer to [1]_.
+
+.. [1] https://docs.kernel.org/PCI/pci-iov-howto.html
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 16/27] pcie_sriov: Make a PCI device with user-created VF ARI-capable
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (14 preceding siblings ...)
2025-05-14 11:50 ` [PULL 15/27] docs: Document composable SR-IOV device Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 17/27] pci-testdev.c: Add membar-backed option for backing membar Michael S. Tsirkin
` (11 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Akihiko Odaki, Yui Washizu, Pasha Tatashin,
Marcel Apfelbaum
From: Akihiko Odaki <akihiko.odaki@daynix.com>
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250314-sriov-v9-9-57dae8ae3ab5@daynix.com>
Tested-by: Yui Washizu <yui.washidu@gmail.com>
Tested-by: Pasha Tatashin <pasha.tatashin@soleen.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/pci/pcie_sriov.h | 7 +++++--
hw/pci/pcie_sriov.c | 8 +++++++-
hw/virtio/virtio-pci.c | 16 ++++++++++------
docs/system/sriov.rst | 3 ++-
4 files changed, 24 insertions(+), 10 deletions(-)
diff --git a/include/hw/pci/pcie_sriov.h b/include/hw/pci/pcie_sriov.h
index f75b8f22ee..aeaa38cf34 100644
--- a/include/hw/pci/pcie_sriov.h
+++ b/include/hw/pci/pcie_sriov.h
@@ -43,12 +43,15 @@ void pcie_sriov_vf_register_bar(PCIDevice *dev, int region_num,
/**
* pcie_sriov_pf_init_from_user_created_vfs() - Initialize PF with user-created
- * VFs.
+ * VFs, adding ARI to PF
* @dev: A PCIe device being realized.
* @offset: The offset of the SR-IOV capability.
* @errp: pointer to Error*, to store an error if it happens.
*
- * Return: The size of added capability. 0 if the user did not create VFs.
+ * Initializes a PF with user-created VFs, adding the ARI extended capability to
+ * the PF. The VFs should call pcie_ari_init() to form an ARI device.
+ *
+ * Return: The size of added capabilities. 0 if the user did not create VFs.
* -1 if failed.
*/
int16_t pcie_sriov_pf_init_from_user_created_vfs(PCIDevice *dev,
diff --git a/hw/pci/pcie_sriov.c b/hw/pci/pcie_sriov.c
index 08f707e847..3ad18744f4 100644
--- a/hw/pci/pcie_sriov.c
+++ b/hw/pci/pcie_sriov.c
@@ -245,6 +245,7 @@ int16_t pcie_sriov_pf_init_from_user_created_vfs(PCIDevice *dev,
PCIDevice **vfs;
BusState *bus = qdev_get_parent_bus(DEVICE(dev));
uint16_t ven_id = pci_get_word(dev->config + PCI_VENDOR_ID);
+ uint16_t size = PCI_EXT_CAP_SRIOV_SIZEOF;
uint16_t vf_dev_id;
uint16_t vf_offset;
uint16_t vf_stride;
@@ -311,6 +312,11 @@ int16_t pcie_sriov_pf_init_from_user_created_vfs(PCIDevice *dev,
return -1;
}
+ if (!pcie_find_capability(dev, PCI_EXT_CAP_ID_ARI)) {
+ pcie_ari_init(dev, offset + size);
+ size += PCI_ARI_SIZEOF;
+ }
+
for (i = 0; i < pf->len; i++) {
vfs[i]->exp.sriov_vf.pf = dev;
vfs[i]->exp.sriov_vf.vf_number = i;
@@ -331,7 +337,7 @@ int16_t pcie_sriov_pf_init_from_user_created_vfs(PCIDevice *dev,
}
}
- return PCI_EXT_CAP_SRIOV_SIZEOF;
+ return size;
}
bool pcie_sriov_register_device(PCIDevice *dev, Error **errp)
diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
index fee65d3645..9b48aa8c3e 100644
--- a/hw/virtio/virtio-pci.c
+++ b/hw/virtio/virtio-pci.c
@@ -2111,12 +2111,16 @@ static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
PCI_BASE_ADDRESS_SPACE_IO, &proxy->bar);
}
- res = pcie_sriov_pf_init_from_user_created_vfs(&proxy->pci_dev,
- proxy->last_pcie_cap_offset,
- errp);
- if (res > 0) {
- proxy->last_pcie_cap_offset += res;
- virtio_add_feature(&vdev->host_features, VIRTIO_F_SR_IOV);
+ if (pci_is_vf(&proxy->pci_dev)) {
+ pcie_ari_init(&proxy->pci_dev, proxy->last_pcie_cap_offset);
+ proxy->last_pcie_cap_offset += PCI_ARI_SIZEOF;
+ } else {
+ res = pcie_sriov_pf_init_from_user_created_vfs(
+ &proxy->pci_dev, proxy->last_pcie_cap_offset, errp);
+ if (res > 0) {
+ proxy->last_pcie_cap_offset += res;
+ virtio_add_feature(&vdev->host_features, VIRTIO_F_SR_IOV);
+ }
}
}
diff --git a/docs/system/sriov.rst b/docs/system/sriov.rst
index a851a66a4b..d12178f3c3 100644
--- a/docs/system/sriov.rst
+++ b/docs/system/sriov.rst
@@ -28,7 +28,8 @@ virtio-net-pci functions to a bus. Below is a command line example:
The VFs specify the paired PF with ``sriov-pf`` property. The PF must be
added after all VFs. It is the user's responsibility to ensure that VFs have
function numbers larger than one of the PF, and that the function numbers
-have a consistent stride.
+have a consistent stride. Both the PF and VFs are ARI-capable so you can have
+255 VFs at maximum.
You may also need to perform additional steps to activate the SR-IOV feature on
your guest. For Linux, refer to [1]_.
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 17/27] pci-testdev.c: Add membar-backed option for backing membar
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (15 preceding siblings ...)
2025-05-14 11:50 ` [PULL 16/27] pcie_sriov: Make a PCI device with user-created VF ARI-capable Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:50 ` [PULL 18/27] system/runstate: add VM state change cb with return value Michael S. Tsirkin
` (10 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Stephen Bates, Marcel Apfelbaum
From: Stephen Bates <sbates@raithlin.com>
The pci-testdev device allows for an optional BAR. We have
historically used this without backing to test that systems and OSes
can accomodate large PCI BARs. However to help test p2pdma operations
it is helpful to add an option to back this BAR with host memory.
We add a membar-backed boolean parameter and when set to true or on we
do a host RAM backing. The default is false which ensures backward
compatability.
Signed-off-by: Stephen Bates <sbates@raithlin.com>
Message-Id: <Z_6JhDtn5PlaDgB_@MKMSTEBATES01.amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/misc/pci-testdev.c | 12 ++++++++++--
1 file changed, 10 insertions(+), 2 deletions(-)
diff --git a/hw/misc/pci-testdev.c b/hw/misc/pci-testdev.c
index 3f6a8bba84..ba71c5069f 100644
--- a/hw/misc/pci-testdev.c
+++ b/hw/misc/pci-testdev.c
@@ -90,6 +90,7 @@ struct PCITestDevState {
int current;
uint64_t membar_size;
+ bool membar_backed;
MemoryRegion membar;
};
@@ -258,8 +259,14 @@ static void pci_testdev_realize(PCIDevice *pci_dev, Error **errp)
pci_register_bar(pci_dev, 1, PCI_BASE_ADDRESS_SPACE_IO, &d->portio);
if (d->membar_size) {
- memory_region_init(&d->membar, OBJECT(d), "pci-testdev-membar",
- d->membar_size);
+ if (d->membar_backed)
+ memory_region_init_ram(&d->membar, OBJECT(d),
+ "pci-testdev-membar-backed",
+ d->membar_size, NULL);
+ else
+ memory_region_init(&d->membar, OBJECT(d),
+ "pci-testdev-membar",
+ d->membar_size);
pci_register_bar(pci_dev, 2,
PCI_BASE_ADDRESS_SPACE_MEMORY |
PCI_BASE_ADDRESS_MEM_PREFETCH |
@@ -321,6 +328,7 @@ static void qdev_pci_testdev_reset(DeviceState *dev)
static const Property pci_testdev_properties[] = {
DEFINE_PROP_SIZE("membar", PCITestDevState, membar_size, 0),
+ DEFINE_PROP_BOOL("membar-backed", PCITestDevState, membar_backed, false),
};
static void pci_testdev_class_init(ObjectClass *klass, const void *data)
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 18/27] system/runstate: add VM state change cb with return value
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (16 preceding siblings ...)
2025-05-14 11:50 ` [PULL 17/27] pci-testdev.c: Add membar-backed option for backing membar Michael S. Tsirkin
@ 2025-05-14 11:50 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 19/27] vhost: return failure if stop virtqueue failed in vhost_dev_stop Michael S. Tsirkin
` (9 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:50 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Haoqian He, Lei Yang, Stefan Hajnoczi, Kevin Wolf,
Hanna Reitz, Paolo Bonzini, Fam Zheng, Alex Williamson,
Cédric Le Goater, Richard Henderson, qemu-block
From: Haoqian He <haoqian.he@smartx.com>
This patch adds the new VM state change cb type `VMChangeStateHandlerWithRet`,
which has return value for `VMChangeStateEntry`.
Thus, we can register a new VM state change cb with return value for device.
Note that `VMChangeStateHandler` and `VMChangeStateHandlerWithRet` are mutually
exclusive and cannot be provided at the same time.
This patch is the pre patch for 'vhost-user: return failure if backend crashes
when live migration', which makes the live migration aware of the loss of
connection with the vhost-user backend and aborts the live migration.
Virtio device will use VMChangeStateHandlerWithRet.
Signed-off-by: Haoqian He <haoqian.he@smartx.com>
Message-Id: <20250416024729.3289157-2-haoqian.he@smartx.com>
Tested-by: Lei Yang <leiyang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/system/runstate.h | 13 +++++++++---
hw/block/virtio-blk.c | 2 +-
hw/core/vm-change-state-handler.c | 18 ++++++++++------
hw/scsi/scsi-bus.c | 2 +-
hw/vfio/migration.c | 2 +-
hw/virtio/virtio.c | 2 +-
system/cpus.c | 8 +++++--
system/runstate.c | 35 ++++++++++++++++++++++++++-----
8 files changed, 62 insertions(+), 20 deletions(-)
diff --git a/include/system/runstate.h b/include/system/runstate.h
index bffc3719d4..fdd5c4a517 100644
--- a/include/system/runstate.h
+++ b/include/system/runstate.h
@@ -12,6 +12,7 @@ bool runstate_needs_reset(void);
void runstate_replay_enable(void);
typedef void VMChangeStateHandler(void *opaque, bool running, RunState state);
+typedef int VMChangeStateHandlerWithRet(void *opaque, bool running, RunState state);
VMChangeStateEntry *qemu_add_vm_change_state_handler(VMChangeStateHandler *cb,
void *opaque);
@@ -20,21 +21,27 @@ VMChangeStateEntry *qemu_add_vm_change_state_handler_prio(
VMChangeStateEntry *
qemu_add_vm_change_state_handler_prio_full(VMChangeStateHandler *cb,
VMChangeStateHandler *prepare_cb,
+ VMChangeStateHandlerWithRet *cb_ret,
void *opaque, int priority);
VMChangeStateEntry *qdev_add_vm_change_state_handler(DeviceState *dev,
VMChangeStateHandler *cb,
+ VMChangeStateHandlerWithRet *cb_ret,
void *opaque);
VMChangeStateEntry *qdev_add_vm_change_state_handler_full(
- DeviceState *dev, VMChangeStateHandler *cb,
- VMChangeStateHandler *prepare_cb, void *opaque);
+ DeviceState *dev, VMChangeStateHandler *cb, VMChangeStateHandler *prepare_cb,
+ VMChangeStateHandlerWithRet *cb_ret, void *opaque);
void qemu_del_vm_change_state_handler(VMChangeStateEntry *e);
/**
* vm_state_notify: Notify the state of the VM
*
* @running: whether the VM is running or not.
* @state: the #RunState of the VM.
+ *
+ * Return the result of the callback which has return value.
+ * If no callback has return value, still return 0 and the
+ * upper layer should not do additional processing.
*/
-void vm_state_notify(bool running, RunState state);
+int vm_state_notify(bool running, RunState state);
static inline bool shutdown_caused_by_guest(ShutdownCause cause)
{
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index b54d01d3a2..ea948d18fd 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -1802,7 +1802,7 @@ static void virtio_blk_device_realize(DeviceState *dev, Error **errp)
* called after ->start_ioeventfd() has already set blk's AioContext.
*/
s->change =
- qdev_add_vm_change_state_handler(dev, virtio_blk_dma_restart_cb, s);
+ qdev_add_vm_change_state_handler(dev, virtio_blk_dma_restart_cb, NULL, s);
blk_ram_registrar_init(&s->blk_ram_registrar, s->blk);
blk_set_dev_ops(s->blk, &virtio_block_ops, s);
diff --git a/hw/core/vm-change-state-handler.c b/hw/core/vm-change-state-handler.c
index 7064995578..99c642b558 100644
--- a/hw/core/vm-change-state-handler.c
+++ b/hw/core/vm-change-state-handler.c
@@ -40,6 +40,7 @@ static int qdev_get_dev_tree_depth(DeviceState *dev)
* qdev_add_vm_change_state_handler:
* @dev: the device that owns this handler
* @cb: the callback function to be invoked
+ * @cb_ret: the callback function with return value to be invoked
* @opaque: user data passed to the callback function
*
* This function works like qemu_add_vm_change_state_handler() except callbacks
@@ -50,25 +51,30 @@ static int qdev_get_dev_tree_depth(DeviceState *dev)
* controller's callback is invoked before the children on its bus when the VM
* starts running. The order is reversed when the VM stops running.
*
+ * Note that the parameter `cb` and `cb_ret` are mutually exclusive.
+ *
* Returns: an entry to be freed with qemu_del_vm_change_state_handler()
*/
VMChangeStateEntry *qdev_add_vm_change_state_handler(DeviceState *dev,
VMChangeStateHandler *cb,
+ VMChangeStateHandlerWithRet *cb_ret,
void *opaque)
{
- return qdev_add_vm_change_state_handler_full(dev, cb, NULL, opaque);
+ assert(!cb || !cb_ret);
+ return qdev_add_vm_change_state_handler_full(dev, cb, NULL, cb_ret, opaque);
}
/*
* Exactly like qdev_add_vm_change_state_handler() but passes a prepare_cb
- * argument too.
+ * and the cb_ret arguments too.
*/
VMChangeStateEntry *qdev_add_vm_change_state_handler_full(
- DeviceState *dev, VMChangeStateHandler *cb,
- VMChangeStateHandler *prepare_cb, void *opaque)
+ DeviceState *dev, VMChangeStateHandler *cb, VMChangeStateHandler *prepare_cb,
+ VMChangeStateHandlerWithRet *cb_ret, void *opaque)
{
int depth = qdev_get_dev_tree_depth(dev);
- return qemu_add_vm_change_state_handler_prio_full(cb, prepare_cb, opaque,
- depth);
+ assert(!cb || !cb_ret);
+ return qemu_add_vm_change_state_handler_prio_full(cb, prepare_cb, cb_ret,
+ opaque, depth);
}
diff --git a/hw/scsi/scsi-bus.c b/hw/scsi/scsi-bus.c
index 70be4a7367..9b12ee7f1c 100644
--- a/hw/scsi/scsi-bus.c
+++ b/hw/scsi/scsi-bus.c
@@ -400,7 +400,7 @@ static void scsi_qdev_realize(DeviceState *qdev, Error **errp)
return;
}
dev->vmsentry = qdev_add_vm_change_state_handler(DEVICE(dev),
- scsi_dma_restart_cb, dev);
+ scsi_dma_restart_cb, NULL, dev);
}
static void scsi_qdev_unrealize(DeviceState *qdev)
diff --git a/hw/vfio/migration.c b/hw/vfio/migration.c
index 1dceab1b19..b76697bd1a 100644
--- a/hw/vfio/migration.c
+++ b/hw/vfio/migration.c
@@ -1016,7 +1016,7 @@ static int vfio_migration_init(VFIODevice *vbasedev)
vfio_vmstate_change_prepare :
NULL;
migration->vm_state = qdev_add_vm_change_state_handler_full(
- vbasedev->dev, vfio_vmstate_change, prepare_cb, vbasedev);
+ vbasedev->dev, vfio_vmstate_change, prepare_cb, NULL, vbasedev);
migration_add_notifier(&migration->migration_state,
vfio_migration_state_notifier);
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 480c2e5036..e3b62522ec 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -3489,7 +3489,7 @@ void virtio_init(VirtIODevice *vdev, uint16_t device_id, size_t config_size)
vdev->config = NULL;
}
vdev->vmstate = qdev_add_vm_change_state_handler(DEVICE(vdev),
- virtio_vmstate_change, vdev);
+ virtio_vmstate_change, NULL, vdev);
vdev->device_endian = virtio_default_endian();
vdev->use_guest_notifier_mask = true;
}
diff --git a/system/cpus.c b/system/cpus.c
index 2cc5f887ab..d16b0dff98 100644
--- a/system/cpus.c
+++ b/system/cpus.c
@@ -299,14 +299,18 @@ static int do_vm_stop(RunState state, bool send_stop)
if (oldstate == RUN_STATE_RUNNING) {
pause_all_vcpus();
}
- vm_state_notify(0, state);
+ ret = vm_state_notify(0, state);
if (send_stop) {
qapi_event_send_stop();
}
}
bdrv_drain_all();
- ret = bdrv_flush_all();
+ /*
+ * Even if vm_state_notify() return failure,
+ * it would be better to flush as before.
+ */
+ ret |= bdrv_flush_all();
trace_vm_stop_flush_all(ret);
return ret;
diff --git a/system/runstate.c b/system/runstate.c
index 272801d307..de74d962bc 100644
--- a/system/runstate.c
+++ b/system/runstate.c
@@ -297,6 +297,7 @@ void qemu_system_vmstop_request(RunState state)
struct VMChangeStateEntry {
VMChangeStateHandler *cb;
VMChangeStateHandler *prepare_cb;
+ VMChangeStateHandlerWithRet *cb_ret;
void *opaque;
QTAILQ_ENTRY(VMChangeStateEntry) entries;
int priority;
@@ -320,14 +321,15 @@ static QTAILQ_HEAD(, VMChangeStateEntry) vm_change_state_head =
VMChangeStateEntry *qemu_add_vm_change_state_handler_prio(
VMChangeStateHandler *cb, void *opaque, int priority)
{
- return qemu_add_vm_change_state_handler_prio_full(cb, NULL, opaque,
- priority);
+ return qemu_add_vm_change_state_handler_prio_full(cb, NULL, NULL,
+ opaque, priority);
}
/**
* qemu_add_vm_change_state_handler_prio_full:
* @cb: the main callback to invoke
* @prepare_cb: a callback to invoke before the main callback
+ * @cb_ret: the main callback to invoke with return value
* @opaque: user data passed to the callbacks
* @priority: low priorities execute first when the vm runs and the reverse is
* true when the vm stops
@@ -344,6 +346,7 @@ VMChangeStateEntry *qemu_add_vm_change_state_handler_prio(
VMChangeStateEntry *
qemu_add_vm_change_state_handler_prio_full(VMChangeStateHandler *cb,
VMChangeStateHandler *prepare_cb,
+ VMChangeStateHandlerWithRet *cb_ret,
void *opaque, int priority)
{
VMChangeStateEntry *e;
@@ -352,6 +355,7 @@ qemu_add_vm_change_state_handler_prio_full(VMChangeStateHandler *cb,
e = g_malloc0(sizeof(*e));
e->cb = cb;
e->prepare_cb = prepare_cb;
+ e->cb_ret = cb_ret;
e->opaque = opaque;
e->priority = priority;
@@ -379,9 +383,10 @@ void qemu_del_vm_change_state_handler(VMChangeStateEntry *e)
g_free(e);
}
-void vm_state_notify(bool running, RunState state)
+int vm_state_notify(bool running, RunState state)
{
VMChangeStateEntry *e, *next;
+ int ret = 0;
trace_vm_state_notify(running, state, RunState_str(state));
@@ -393,7 +398,17 @@ void vm_state_notify(bool running, RunState state)
}
QTAILQ_FOREACH_SAFE(e, &vm_change_state_head, entries, next) {
- e->cb(e->opaque, running, state);
+ if (e->cb) {
+ e->cb(e->opaque, running, state);
+ } else if (e->cb_ret) {
+ /*
+ * Here ignore the return value of cb_ret because
+ * we only care about the stopping the device during
+ * the VM live migration to indicate whether the
+ * connection between qemu and backend is normal.
+ */
+ e->cb_ret(e->opaque, running, state);
+ }
}
} else {
QTAILQ_FOREACH_REVERSE_SAFE(e, &vm_change_state_head, entries, next) {
@@ -403,9 +418,19 @@ void vm_state_notify(bool running, RunState state)
}
QTAILQ_FOREACH_REVERSE_SAFE(e, &vm_change_state_head, entries, next) {
- e->cb(e->opaque, running, state);
+ if (e->cb) {
+ e->cb(e->opaque, running, state);
+ } else if (e->cb_ret) {
+ /*
+ * We should execute all registered callbacks even if
+ * one of them returns failure, otherwise, some cleanup
+ * work of the device will be skipped.
+ */
+ ret |= e->cb_ret(e->opaque, running, state);
+ }
}
}
+ return ret;
}
static ShutdownCause reset_requested;
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 19/27] vhost: return failure if stop virtqueue failed in vhost_dev_stop
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (17 preceding siblings ...)
2025-05-14 11:50 ` [PULL 18/27] system/runstate: add VM state change cb with return value Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 20/27] vhost-user: return failure if backend crash when live migration Michael S. Tsirkin
` (8 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Haoqian He, Lei Yang, Stefano Garzarella
From: Haoqian He <haoqian.he@smartx.com>
This patch captures the error of vhost_virtqueue_stop() in vhost_dev_stop()
and returns the error upward.
Specifically, if QEMU is disconnected from the vhost backend, some actions
in vhost_dev_stop() will fail, such as sending vhost-user messages to the
backend (GET_VRING_BASE, SET_VRING_ENABLE) and vhost_reset_status.
Considering that both set_vring_enable and vhost_reset_status require setting
the specific virtio feature bit, we can capture vhost_virtqueue_stop()'s
error to indicate that QEMU has lost connection with the backend.
This patch is the pre patch for 'vhost-user: return failure if backend crashes
when live migration', which makes the live migration aware of the loss of
connection with the vhost-user backend and aborts the live migration.
Signed-off-by: Haoqian He <haoqian.he@smartx.com>
Message-Id: <20250416024729.3289157-3-haoqian.he@smartx.com>
Tested-by: Lei Yang <leiyang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/virtio/vhost.h | 8 +++++---
hw/virtio/vhost.c | 23 +++++++++++++----------
2 files changed, 18 insertions(+), 13 deletions(-)
diff --git a/include/hw/virtio/vhost.h b/include/hw/virtio/vhost.h
index bb4b58e115..38800a7156 100644
--- a/include/hw/virtio/vhost.h
+++ b/include/hw/virtio/vhost.h
@@ -232,8 +232,10 @@ int vhost_dev_start(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings);
* Stop the vhost device. After the device is stopped the notifiers
* can be disabled (@vhost_dev_disable_notifiers) and the device can
* be torn down (@vhost_dev_cleanup).
+ *
+ * Return: 0 on success, != 0 on error when stopping dev.
*/
-void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings);
+int vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings);
/**
* DOC: vhost device configuration handling
@@ -333,8 +335,8 @@ int vhost_device_iotlb_miss(struct vhost_dev *dev, uint64_t iova, int write);
int vhost_virtqueue_start(struct vhost_dev *dev, struct VirtIODevice *vdev,
struct vhost_virtqueue *vq, unsigned idx);
-void vhost_virtqueue_stop(struct vhost_dev *dev, struct VirtIODevice *vdev,
- struct vhost_virtqueue *vq, unsigned idx);
+int vhost_virtqueue_stop(struct vhost_dev *dev, struct VirtIODevice *vdev,
+ struct vhost_virtqueue *vq, unsigned idx);
void vhost_dev_reset_inflight(struct vhost_inflight *inflight);
void vhost_dev_free_inflight(struct vhost_inflight *inflight);
diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
index 4cae7c1664..fc43853704 100644
--- a/hw/virtio/vhost.c
+++ b/hw/virtio/vhost.c
@@ -1367,10 +1367,10 @@ fail_alloc_desc:
return r;
}
-void vhost_virtqueue_stop(struct vhost_dev *dev,
- struct VirtIODevice *vdev,
- struct vhost_virtqueue *vq,
- unsigned idx)
+int vhost_virtqueue_stop(struct vhost_dev *dev,
+ struct VirtIODevice *vdev,
+ struct vhost_virtqueue *vq,
+ unsigned idx)
{
int vhost_vq_index = dev->vhost_ops->vhost_get_vq_index(dev, idx);
struct vhost_vring_state state = {
@@ -1380,7 +1380,7 @@ void vhost_virtqueue_stop(struct vhost_dev *dev,
if (virtio_queue_get_desc_addr(vdev, idx) == 0) {
/* Don't stop the virtqueue which might have not been started */
- return;
+ return 0;
}
r = dev->vhost_ops->vhost_get_vring_base(dev, &state);
@@ -1411,6 +1411,7 @@ void vhost_virtqueue_stop(struct vhost_dev *dev,
0, virtio_queue_get_avail_size(vdev, idx));
vhost_memory_unmap(dev, vq->desc, virtio_queue_get_desc_size(vdev, idx),
0, virtio_queue_get_desc_size(vdev, idx));
+ return r;
}
static int vhost_virtqueue_set_busyloop_timeout(struct vhost_dev *dev,
@@ -2135,9 +2136,10 @@ fail_features:
}
/* Host notifiers must be enabled at this point. */
-void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings)
+int vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings)
{
int i;
+ int rc = 0;
/* should only be called after backend is connected */
assert(hdev->vhost_ops);
@@ -2156,10 +2158,10 @@ void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings)
vhost_dev_set_vring_enable(hdev, false);
}
for (i = 0; i < hdev->nvqs; ++i) {
- vhost_virtqueue_stop(hdev,
- vdev,
- hdev->vqs + i,
- hdev->vq_index + i);
+ rc |= vhost_virtqueue_stop(hdev,
+ vdev,
+ hdev->vqs + i,
+ hdev->vq_index + i);
}
if (hdev->vhost_ops->vhost_reset_status) {
hdev->vhost_ops->vhost_reset_status(hdev);
@@ -2176,6 +2178,7 @@ void vhost_dev_stop(struct vhost_dev *hdev, VirtIODevice *vdev, bool vrings)
hdev->started = false;
vdev->vhost_started = false;
hdev->vdev = NULL;
+ return rc;
}
int vhost_net_set_backend(struct vhost_dev *hdev,
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 20/27] vhost-user: return failure if backend crash when live migration
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (18 preceding siblings ...)
2025-05-14 11:51 ` [PULL 19/27] vhost: return failure if stop virtqueue failed in vhost_dev_stop Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 21/27] vhost-scsi: support VIRTIO_SCSI_F_HOTPLUG Michael S. Tsirkin
` (7 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Haoqian He, Lei Yang, Stefano Garzarella,
Raphael Norwitz, Kevin Wolf, Hanna Reitz, Stefan Hajnoczi,
Laurent Vivier, Amit Shah, Marc-André Lureau, Paolo Bonzini,
Gerd Hoffmann, Jason Wang, Fam Zheng, Alex Bennée, mzamazal,
David Hildenbrand, Gonglei (Arei), Eric Auger, qemu-block,
virtio-fs
From: Haoqian He <haoqian.he@smartx.com>
Live migration should be terminated if the vhost-user backend crashes
before the migration completes.
Specifically, since the vhost device will be stopped when VM is stopped
before the end of the live migration, in current implementation if the
backend crashes, vhost-user device set_status() won't return failure,
live migration won't perceive the disconnection between QEMU and the
backend.
When the VM is migrated to the destination, the inflight IO will be
resubmitted, and if the IO was completed out of order before, it will
cause IO error.
To fix this issue:
1. Add the return value to set_status() for VirtioDeviceClass.
a. For the vhost-user device, return failure when the backend crashes.
b. For other virtio devices, always return 0.
2. Return failure if vhost_dev_stop() failed for vhost-user device.
If QEMU loses connection with the vhost-user backend, virtio set_status()
can return failure to the upper layer, migration_completion() can handle
the error, terminate the live migration, and restore the VM, so that
inflight IO can be completed normally.
Signed-off-by: Haoqian He <haoqian.he@smartx.com>
Message-Id: <20250416024729.3289157-4-haoqian.he@smartx.com>
Tested-by: Lei Yang <leiyang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
include/hw/virtio/vhost-scsi-common.h | 2 +-
include/hw/virtio/vhost-vsock-common.h | 2 +-
include/hw/virtio/virtio.h | 2 +-
include/system/vhost-user-backend.h | 2 +-
backends/vhost-user.c | 20 +++++++++----------
hw/block/vhost-user-blk.c | 27 ++++++++++++++------------
hw/block/virtio-blk.c | 5 +++--
hw/char/virtio-serial-bus.c | 3 ++-
hw/display/vhost-user-gpu.c | 12 +++++++++---
hw/input/virtio-input.c | 3 ++-
hw/net/virtio-net.c | 3 ++-
hw/scsi/vhost-scsi-common.c | 13 +++++++------
hw/scsi/vhost-scsi.c | 5 +++--
hw/scsi/vhost-user-scsi.c | 18 ++++++++++-------
hw/virtio/vdpa-dev.c | 5 +++--
hw/virtio/vhost-user-base.c | 23 +++++++++++++---------
hw/virtio/vhost-user-fs.c | 23 +++++++++++++---------
hw/virtio/vhost-user-scmi.c | 27 +++++++++++++++-----------
hw/virtio/vhost-user-vsock.c | 15 +++++++++-----
hw/virtio/vhost-vsock-common.c | 12 ++++++------
hw/virtio/vhost-vsock.c | 11 ++++++-----
hw/virtio/virtio-balloon.c | 3 ++-
hw/virtio/virtio-crypto.c | 3 ++-
hw/virtio/virtio-iommu.c | 3 ++-
hw/virtio/virtio-rng.c | 5 +++--
hw/virtio/virtio.c | 22 ++++++++++++++-------
26 files changed, 160 insertions(+), 109 deletions(-)
diff --git a/include/hw/virtio/vhost-scsi-common.h b/include/hw/virtio/vhost-scsi-common.h
index c5d2c09455..d54d9c916f 100644
--- a/include/hw/virtio/vhost-scsi-common.h
+++ b/include/hw/virtio/vhost-scsi-common.h
@@ -40,7 +40,7 @@ struct VHostSCSICommon {
};
int vhost_scsi_common_start(VHostSCSICommon *vsc, Error **errp);
-void vhost_scsi_common_stop(VHostSCSICommon *vsc);
+int vhost_scsi_common_stop(VHostSCSICommon *vsc);
char *vhost_scsi_common_get_fw_dev_path(FWPathProvider *p, BusState *bus,
DeviceState *dev);
void vhost_scsi_common_set_config(VirtIODevice *vdev, const uint8_t *config);
diff --git a/include/hw/virtio/vhost-vsock-common.h b/include/hw/virtio/vhost-vsock-common.h
index 75a74e8a99..01bf6062af 100644
--- a/include/hw/virtio/vhost-vsock-common.h
+++ b/include/hw/virtio/vhost-vsock-common.h
@@ -42,7 +42,7 @@ struct VHostVSockCommon {
};
int vhost_vsock_common_start(VirtIODevice *vdev);
-void vhost_vsock_common_stop(VirtIODevice *vdev);
+int vhost_vsock_common_stop(VirtIODevice *vdev);
int vhost_vsock_common_pre_save(void *opaque);
int vhost_vsock_common_post_load(void *opaque, int version_id);
void vhost_vsock_common_realize(VirtIODevice *vdev);
diff --git a/include/hw/virtio/virtio.h b/include/hw/virtio/virtio.h
index 7e0c471ea4..214d4a77e9 100644
--- a/include/hw/virtio/virtio.h
+++ b/include/hw/virtio/virtio.h
@@ -186,7 +186,7 @@ struct VirtioDeviceClass {
void (*get_config)(VirtIODevice *vdev, uint8_t *config);
void (*set_config)(VirtIODevice *vdev, const uint8_t *config);
void (*reset)(VirtIODevice *vdev);
- void (*set_status)(VirtIODevice *vdev, uint8_t val);
+ int (*set_status)(VirtIODevice *vdev, uint8_t val);
/* Device must validate queue_index. */
void (*queue_reset)(VirtIODevice *vdev, uint32_t queue_index);
/* Device must validate queue_index. */
diff --git a/include/system/vhost-user-backend.h b/include/system/vhost-user-backend.h
index 5ed953cd53..5634ebdb2e 100644
--- a/include/system/vhost-user-backend.h
+++ b/include/system/vhost-user-backend.h
@@ -43,6 +43,6 @@ struct VhostUserBackend {
int vhost_user_backend_dev_init(VhostUserBackend *b, VirtIODevice *vdev,
unsigned nvqs, Error **errp);
void vhost_user_backend_start(VhostUserBackend *b);
-void vhost_user_backend_stop(VhostUserBackend *b);
+int vhost_user_backend_stop(VhostUserBackend *b);
#endif
diff --git a/backends/vhost-user.c b/backends/vhost-user.c
index 94274a619d..42845329e7 100644
--- a/backends/vhost-user.c
+++ b/backends/vhost-user.c
@@ -97,30 +97,28 @@ err_host_notifiers:
vhost_dev_disable_notifiers(&b->dev, b->vdev);
}
-void
+int
vhost_user_backend_stop(VhostUserBackend *b)
{
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(b->vdev)));
VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
- int ret = 0;
+ int ret;
if (!b->started) {
- return;
+ return 0;
}
- vhost_dev_stop(&b->dev, b->vdev, true);
+ ret = vhost_dev_stop(&b->dev, b->vdev, true);
- if (k->set_guest_notifiers) {
- ret = k->set_guest_notifiers(qbus->parent,
- b->dev.nvqs, false);
- if (ret < 0) {
- error_report("vhost guest notifier cleanup failed: %d", ret);
- }
+ if (k->set_guest_notifiers &&
+ k->set_guest_notifiers(qbus->parent, b->dev.nvqs, false) < 0) {
+ error_report("vhost guest notifier cleanup failed: %d", ret);
+ return -1;
}
- assert(ret >= 0);
vhost_dev_disable_notifiers(&b->dev, b->vdev);
b->started = false;
+ return ret;
}
static void set_chardev(Object *obj, const char *value, Error **errp)
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c
index 4bb5ed299e..0eebbcd80d 100644
--- a/hw/block/vhost-user-blk.c
+++ b/hw/block/vhost-user-blk.c
@@ -204,7 +204,7 @@ err_host_notifiers:
return ret;
}
-static void vhost_user_blk_stop(VirtIODevice *vdev)
+static int vhost_user_blk_stop(VirtIODevice *vdev)
{
VHostUserBlk *s = VHOST_USER_BLK(vdev);
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
@@ -212,26 +212,26 @@ static void vhost_user_blk_stop(VirtIODevice *vdev)
int ret;
if (!s->started_vu) {
- return;
+ return 0;
}
s->started_vu = false;
if (!k->set_guest_notifiers) {
- return;
+ return 0;
}
- vhost_dev_stop(&s->dev, vdev, true);
+ ret = vhost_dev_stop(&s->dev, vdev, true);
- ret = k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false);
- if (ret < 0) {
+ if (k->set_guest_notifiers(qbus->parent, s->dev.nvqs, false) < 0) {
error_report("vhost guest notifier cleanup failed: %d", ret);
- return;
+ return -1;
}
vhost_dev_disable_notifiers(&s->dev, vdev);
+ return ret;
}
-static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
+static int vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserBlk *s = VHOST_USER_BLK(vdev);
bool should_start = virtio_device_should_start(vdev, status);
@@ -239,11 +239,11 @@ static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
int ret;
if (!s->connected) {
- return;
+ return -1;
}
if (vhost_dev_is_started(&s->dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
@@ -253,9 +253,12 @@ static void vhost_user_blk_set_status(VirtIODevice *vdev, uint8_t status)
qemu_chr_fe_disconnect(&s->chardev);
}
} else {
- vhost_user_blk_stop(vdev);
+ ret = vhost_user_blk_stop(vdev);
+ if (ret < 0) {
+ return ret;
+ }
}
-
+ return 0;
}
static uint64_t vhost_user_blk_get_features(VirtIODevice *vdev,
diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c
index ea948d18fd..9bab2716c1 100644
--- a/hw/block/virtio-blk.c
+++ b/hw/block/virtio-blk.c
@@ -1270,7 +1270,7 @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features,
return features;
}
-static void virtio_blk_set_status(VirtIODevice *vdev, uint8_t status)
+static int virtio_blk_set_status(VirtIODevice *vdev, uint8_t status)
{
VirtIOBlock *s = VIRTIO_BLK(vdev);
@@ -1279,7 +1279,7 @@ static void virtio_blk_set_status(VirtIODevice *vdev, uint8_t status)
}
if (!(status & VIRTIO_CONFIG_S_DRIVER_OK)) {
- return;
+ return 0;
}
/* A guest that supports VIRTIO_BLK_F_CONFIG_WCE must be able to send
@@ -1302,6 +1302,7 @@ static void virtio_blk_set_status(VirtIODevice *vdev, uint8_t status)
virtio_vdev_has_feature(vdev,
VIRTIO_BLK_F_WCE));
}
+ return 0;
}
static void virtio_blk_save_device(VirtIODevice *vdev, QEMUFile *f)
diff --git a/hw/char/virtio-serial-bus.c b/hw/char/virtio-serial-bus.c
index eb79f5258b..673c50f0be 100644
--- a/hw/char/virtio-serial-bus.c
+++ b/hw/char/virtio-serial-bus.c
@@ -622,7 +622,7 @@ static void guest_reset(VirtIOSerial *vser)
}
}
-static void set_status(VirtIODevice *vdev, uint8_t status)
+static int set_status(VirtIODevice *vdev, uint8_t status)
{
VirtIOSerial *vser;
VirtIOSerialPort *port;
@@ -650,6 +650,7 @@ static void set_status(VirtIODevice *vdev, uint8_t status)
vsc->enable_backend(port, vdev->vm_running);
}
}
+ return 0;
}
static void vser_reset(VirtIODevice *vdev)
diff --git a/hw/display/vhost-user-gpu.c b/hw/display/vhost-user-gpu.c
index 43d4c08a2e..9fc6bbcd2c 100644
--- a/hw/display/vhost-user-gpu.c
+++ b/hw/display/vhost-user-gpu.c
@@ -516,7 +516,7 @@ vhost_user_gpu_set_config(VirtIODevice *vdev,
}
}
-static void
+static int
vhost_user_gpu_set_status(VirtIODevice *vdev, uint8_t val)
{
VhostUserGPU *g = VHOST_USER_GPU(vdev);
@@ -525,18 +525,24 @@ vhost_user_gpu_set_status(VirtIODevice *vdev, uint8_t val)
if (val & VIRTIO_CONFIG_S_DRIVER_OK && vdev->vm_running) {
if (!vhost_user_gpu_do_set_socket(g, &err)) {
error_report_err(err);
- return;
+ return 0;
}
vhost_user_backend_start(g->vhost);
} else {
+ int ret;
+
/* unblock any wait and stop processing */
if (g->vhost_gpu_fd != -1) {
vhost_user_gpu_update_blocked(g, true);
qemu_chr_fe_deinit(&g->vhost_chr, true);
g->vhost_gpu_fd = -1;
}
- vhost_user_backend_stop(g->vhost);
+ ret = vhost_user_backend_stop(g->vhost);
+ if (ret < 0) {
+ return ret;
+ }
}
+ return 0;
}
static bool
diff --git a/hw/input/virtio-input.c b/hw/input/virtio-input.c
index 1818cbddc7..a3f554f211 100644
--- a/hw/input/virtio-input.c
+++ b/hw/input/virtio-input.c
@@ -189,7 +189,7 @@ static uint64_t virtio_input_get_features(VirtIODevice *vdev, uint64_t f,
return f;
}
-static void virtio_input_set_status(VirtIODevice *vdev, uint8_t val)
+static int virtio_input_set_status(VirtIODevice *vdev, uint8_t val)
{
VirtIOInputClass *vic = VIRTIO_INPUT_GET_CLASS(vdev);
VirtIOInput *vinput = VIRTIO_INPUT(vdev);
@@ -202,6 +202,7 @@ static void virtio_input_set_status(VirtIODevice *vdev, uint8_t val)
}
}
}
+ return 0;
}
static void virtio_input_reset(VirtIODevice *vdev)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index 2de037c273..221252e00a 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -382,7 +382,7 @@ static void virtio_net_drop_tx_queue_data(VirtIODevice *vdev, VirtQueue *vq)
}
}
-static void virtio_net_set_status(struct VirtIODevice *vdev, uint8_t status)
+static int virtio_net_set_status(struct VirtIODevice *vdev, uint8_t status)
{
VirtIONet *n = VIRTIO_NET(vdev);
VirtIONetQueue *q;
@@ -437,6 +437,7 @@ static void virtio_net_set_status(struct VirtIODevice *vdev, uint8_t status)
}
}
}
+ return 0;
}
static void virtio_net_set_link_status(NetClientState *nc)
diff --git a/hw/scsi/vhost-scsi-common.c b/hw/scsi/vhost-scsi-common.c
index 4c8637045d..43525ba46d 100644
--- a/hw/scsi/vhost-scsi-common.c
+++ b/hw/scsi/vhost-scsi-common.c
@@ -101,24 +101,25 @@ err_host_notifiers:
return ret;
}
-void vhost_scsi_common_stop(VHostSCSICommon *vsc)
+int vhost_scsi_common_stop(VHostSCSICommon *vsc)
{
VirtIODevice *vdev = VIRTIO_DEVICE(vsc);
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
VirtioBusClass *k = VIRTIO_BUS_GET_CLASS(qbus);
int ret = 0;
- vhost_dev_stop(&vsc->dev, vdev, true);
+ ret = vhost_dev_stop(&vsc->dev, vdev, true);
if (k->set_guest_notifiers) {
- ret = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
- if (ret < 0) {
- error_report("vhost guest notifier cleanup failed: %d", ret);
+ int r = k->set_guest_notifiers(qbus->parent, vsc->dev.nvqs, false);
+ if (r < 0) {
+ error_report("vhost guest notifier cleanup failed: %d", ret);
+ return r;
}
}
- assert(ret >= 0);
vhost_dev_disable_notifiers(&vsc->dev, vdev);
+ return ret;
}
uint64_t vhost_scsi_common_get_features(VirtIODevice *vdev, uint64_t features,
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index 10fde8eee0..dd4250ebe8 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -114,7 +114,7 @@ static void vhost_scsi_stop(VHostSCSI *s)
vhost_scsi_common_stop(vsc);
}
-static void vhost_scsi_set_status(VirtIODevice *vdev, uint8_t val)
+static int vhost_scsi_set_status(VirtIODevice *vdev, uint8_t val)
{
VHostSCSI *s = VHOST_SCSI(vdev);
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
@@ -125,7 +125,7 @@ static void vhost_scsi_set_status(VirtIODevice *vdev, uint8_t val)
}
if (vhost_dev_is_started(&vsc->dev) == start) {
- return;
+ return 0;
}
if (start) {
@@ -139,6 +139,7 @@ static void vhost_scsi_set_status(VirtIODevice *vdev, uint8_t val)
} else {
vhost_scsi_stop(s);
}
+ return 0;
}
static void vhost_dummy_handle_output(VirtIODevice *vdev, VirtQueue *vq)
diff --git a/hw/scsi/vhost-user-scsi.c b/hw/scsi/vhost-user-scsi.c
index 8298e8cc6d..25f2d894e7 100644
--- a/hw/scsi/vhost-user-scsi.c
+++ b/hw/scsi/vhost-user-scsi.c
@@ -52,19 +52,19 @@ static int vhost_user_scsi_start(VHostUserSCSI *s, Error **errp)
return ret;
}
-static void vhost_user_scsi_stop(VHostUserSCSI *s)
+static int vhost_user_scsi_stop(VHostUserSCSI *s)
{
VHostSCSICommon *vsc = VHOST_SCSI_COMMON(s);
if (!s->started_vu) {
- return;
+ return 0;
}
s->started_vu = false;
- vhost_scsi_common_stop(vsc);
+ return vhost_scsi_common_stop(vsc);
}
-static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
+static int vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserSCSI *s = (VHostUserSCSI *)vdev;
DeviceState *dev = DEVICE(vdev);
@@ -75,11 +75,11 @@ static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
int ret;
if (!s->connected) {
- return;
+ return -1;
}
if (vhost_dev_is_started(&vsc->dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
@@ -91,8 +91,12 @@ static void vhost_user_scsi_set_status(VirtIODevice *vdev, uint8_t status)
qemu_chr_fe_disconnect(&vs->conf.chardev);
}
} else {
- vhost_user_scsi_stop(s);
+ ret = vhost_user_scsi_stop(s);
+ if (ret) {
+ return ret;
+ }
}
+ return 0;
}
static void vhost_user_scsi_handle_output(VirtIODevice *vdev, VirtQueue *vq)
diff --git a/hw/virtio/vdpa-dev.c b/hw/virtio/vdpa-dev.c
index dd8837ce4e..d1da40afc8 100644
--- a/hw/virtio/vdpa-dev.c
+++ b/hw/virtio/vdpa-dev.c
@@ -312,7 +312,7 @@ static void vhost_vdpa_device_stop(VirtIODevice *vdev)
vhost_dev_disable_notifiers(&s->dev, vdev);
}
-static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
+static int vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
{
VhostVdpaDevice *s = VHOST_VDPA_DEVICE(vdev);
bool should_start = virtio_device_started(vdev, status);
@@ -324,7 +324,7 @@ static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
}
if (s->started == should_start) {
- return;
+ return 0;
}
if (should_start) {
@@ -335,6 +335,7 @@ static void vhost_vdpa_device_set_status(VirtIODevice *vdev, uint8_t status)
} else {
vhost_vdpa_device_stop(vdev);
}
+ return 0;
}
static const Property vhost_vdpa_device_properties[] = {
diff --git a/hw/virtio/vhost-user-base.c b/hw/virtio/vhost-user-base.c
index 77143320a2..ff67a020b4 100644
--- a/hw/virtio/vhost-user-base.c
+++ b/hw/virtio/vhost-user-base.c
@@ -66,7 +66,7 @@ err_host_notifiers:
vhost_dev_disable_notifiers(&vub->vhost_dev, vdev);
}
-static void vub_stop(VirtIODevice *vdev)
+static int vub_stop(VirtIODevice *vdev)
{
VHostUserBase *vub = VHOST_USER_BASE(vdev);
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
@@ -74,34 +74,39 @@ static void vub_stop(VirtIODevice *vdev)
int ret;
if (!k->set_guest_notifiers) {
- return;
+ return 0;
}
- vhost_dev_stop(&vub->vhost_dev, vdev, true);
+ ret = vhost_dev_stop(&vub->vhost_dev, vdev, true);
- ret = k->set_guest_notifiers(qbus->parent, vub->vhost_dev.nvqs, false);
- if (ret < 0) {
+ if (k->set_guest_notifiers(qbus->parent, vub->vhost_dev.nvqs, false) < 0) {
error_report("vhost guest notifier cleanup failed: %d", ret);
- return;
+ return -1;
}
vhost_dev_disable_notifiers(&vub->vhost_dev, vdev);
+ return ret;
}
-static void vub_set_status(VirtIODevice *vdev, uint8_t status)
+static int vub_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserBase *vub = VHOST_USER_BASE(vdev);
bool should_start = virtio_device_should_start(vdev, status);
if (vhost_dev_is_started(&vub->vhost_dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
vub_start(vdev);
} else {
- vub_stop(vdev);
+ int ret;
+ ret = vub_stop(vdev);
+ if (ret < 0) {
+ return ret;
+ }
}
+ return 0;
}
/*
diff --git a/hw/virtio/vhost-user-fs.c b/hw/virtio/vhost-user-fs.c
index f6d1fc8804..e77c69eb12 100644
--- a/hw/virtio/vhost-user-fs.c
+++ b/hw/virtio/vhost-user-fs.c
@@ -100,7 +100,7 @@ err_host_notifiers:
vhost_dev_disable_notifiers(&fs->vhost_dev, vdev);
}
-static void vuf_stop(VirtIODevice *vdev)
+static int vuf_stop(VirtIODevice *vdev)
{
VHostUserFS *fs = VHOST_USER_FS(vdev);
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
@@ -108,34 +108,39 @@ static void vuf_stop(VirtIODevice *vdev)
int ret;
if (!k->set_guest_notifiers) {
- return;
+ return 0;
}
- vhost_dev_stop(&fs->vhost_dev, vdev, true);
+ ret = vhost_dev_stop(&fs->vhost_dev, vdev, true);
- ret = k->set_guest_notifiers(qbus->parent, fs->vhost_dev.nvqs, false);
- if (ret < 0) {
+ if (k->set_guest_notifiers(qbus->parent, fs->vhost_dev.nvqs, false) < 0) {
error_report("vhost guest notifier cleanup failed: %d", ret);
- return;
+ return -1;
}
vhost_dev_disable_notifiers(&fs->vhost_dev, vdev);
+ return ret;
}
-static void vuf_set_status(VirtIODevice *vdev, uint8_t status)
+static int vuf_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserFS *fs = VHOST_USER_FS(vdev);
bool should_start = virtio_device_should_start(vdev, status);
if (vhost_dev_is_started(&fs->vhost_dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
vuf_start(vdev);
} else {
- vuf_stop(vdev);
+ int ret;
+ ret = vuf_stop(vdev);
+ if (ret < 0) {
+ return ret;
+ }
}
+ return 0;
}
static uint64_t vuf_get_features(VirtIODevice *vdev,
diff --git a/hw/virtio/vhost-user-scmi.c b/hw/virtio/vhost-user-scmi.c
index 7a0f622181..f9264c4374 100644
--- a/hw/virtio/vhost-user-scmi.c
+++ b/hw/virtio/vhost-user-scmi.c
@@ -83,7 +83,7 @@ err_host_notifiers:
return ret;
}
-static void vu_scmi_stop(VirtIODevice *vdev)
+static int vu_scmi_stop(VirtIODevice *vdev)
{
VHostUserSCMI *scmi = VHOST_USER_SCMI(vdev);
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
@@ -93,41 +93,46 @@ static void vu_scmi_stop(VirtIODevice *vdev)
/* vhost_dev_is_started() check in the callers is not fully reliable. */
if (!scmi->started_vu) {
- return;
+ return 0;
}
scmi->started_vu = false;
if (!k->set_guest_notifiers) {
- return;
+ return 0;
}
- vhost_dev_stop(vhost_dev, vdev, true);
+ ret = vhost_dev_stop(vhost_dev, vdev, true);
- ret = k->set_guest_notifiers(qbus->parent, vhost_dev->nvqs, false);
- if (ret < 0) {
+ if (k->set_guest_notifiers(qbus->parent, vhost_dev->nvqs, false) < 0) {
error_report("vhost guest notifier cleanup failed: %d", ret);
- return;
+ return -1;
}
vhost_dev_disable_notifiers(vhost_dev, vdev);
+ return ret;
}
-static void vu_scmi_set_status(VirtIODevice *vdev, uint8_t status)
+static int vu_scmi_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostUserSCMI *scmi = VHOST_USER_SCMI(vdev);
bool should_start = virtio_device_should_start(vdev, status);
if (!scmi->connected) {
- return;
+ return -1;
}
if (vhost_dev_is_started(&scmi->vhost_dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
vu_scmi_start(vdev);
} else {
- vu_scmi_stop(vdev);
+ int ret;
+ ret = vu_scmi_stop(vdev);
+ if (ret < 0) {
+ return ret;
+ }
}
+ return 0;
}
static uint64_t vu_scmi_get_features(VirtIODevice *vdev, uint64_t features,
diff --git a/hw/virtio/vhost-user-vsock.c b/hw/virtio/vhost-user-vsock.c
index 2776792f59..993c287348 100644
--- a/hw/virtio/vhost-user-vsock.c
+++ b/hw/virtio/vhost-user-vsock.c
@@ -54,23 +54,28 @@ const VhostDevConfigOps vsock_ops = {
.vhost_dev_config_notifier = vuv_handle_config_change,
};
-static void vuv_set_status(VirtIODevice *vdev, uint8_t status)
+static int vuv_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostVSockCommon *vvc = VHOST_VSOCK_COMMON(vdev);
bool should_start = virtio_device_should_start(vdev, status);
+ int ret;
if (vhost_dev_is_started(&vvc->vhost_dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
- int ret = vhost_vsock_common_start(vdev);
+ ret = vhost_vsock_common_start(vdev);
if (ret < 0) {
- return;
+ return ret;
}
} else {
- vhost_vsock_common_stop(vdev);
+ ret = vhost_vsock_common_stop(vdev);
+ if (ret < 0) {
+ return ret;
+ }
}
+ return 0;
}
static uint64_t vuv_get_features(VirtIODevice *vdev,
diff --git a/hw/virtio/vhost-vsock-common.c b/hw/virtio/vhost-vsock-common.c
index 4b4fbb45cc..c6c44d8989 100644
--- a/hw/virtio/vhost-vsock-common.c
+++ b/hw/virtio/vhost-vsock-common.c
@@ -95,7 +95,7 @@ err_host_notifiers:
return ret;
}
-void vhost_vsock_common_stop(VirtIODevice *vdev)
+int vhost_vsock_common_stop(VirtIODevice *vdev)
{
VHostVSockCommon *vvc = VHOST_VSOCK_COMMON(vdev);
BusState *qbus = BUS(qdev_get_parent_bus(DEVICE(vdev)));
@@ -103,18 +103,18 @@ void vhost_vsock_common_stop(VirtIODevice *vdev)
int ret;
if (!k->set_guest_notifiers) {
- return;
+ return 0;
}
- vhost_dev_stop(&vvc->vhost_dev, vdev, true);
+ ret = vhost_dev_stop(&vvc->vhost_dev, vdev, true);
- ret = k->set_guest_notifiers(qbus->parent, vvc->vhost_dev.nvqs, false);
- if (ret < 0) {
+ if (k->set_guest_notifiers(qbus->parent, vvc->vhost_dev.nvqs, false) < 0) {
error_report("vhost guest notifier cleanup failed: %d", ret);
- return;
+ return -1;
}
vhost_dev_disable_notifiers(&vvc->vhost_dev, vdev);
+ return ret;
}
diff --git a/hw/virtio/vhost-vsock.c b/hw/virtio/vhost-vsock.c
index b73dc723c2..6e4088831f 100644
--- a/hw/virtio/vhost-vsock.c
+++ b/hw/virtio/vhost-vsock.c
@@ -67,37 +67,38 @@ static int vhost_vsock_set_running(VirtIODevice *vdev, int start)
}
-static void vhost_vsock_set_status(VirtIODevice *vdev, uint8_t status)
+static int vhost_vsock_set_status(VirtIODevice *vdev, uint8_t status)
{
VHostVSockCommon *vvc = VHOST_VSOCK_COMMON(vdev);
bool should_start = virtio_device_should_start(vdev, status);
int ret;
if (vhost_dev_is_started(&vvc->vhost_dev) == should_start) {
- return;
+ return 0;
}
if (should_start) {
ret = vhost_vsock_common_start(vdev);
if (ret < 0) {
- return;
+ return 0;
}
ret = vhost_vsock_set_running(vdev, 1);
if (ret < 0) {
vhost_vsock_common_stop(vdev);
error_report("Error starting vhost vsock: %d", -ret);
- return;
+ return 0;
}
} else {
ret = vhost_vsock_set_running(vdev, 0);
if (ret < 0) {
error_report("vhost vsock set running failed: %d", ret);
- return;
+ return 0;
}
vhost_vsock_common_stop(vdev);
}
+ return 0;
}
static uint64_t vhost_vsock_get_features(VirtIODevice *vdev,
diff --git a/hw/virtio/virtio-balloon.c b/hw/virtio/virtio-balloon.c
index 91510ec2e2..db787d00b3 100644
--- a/hw/virtio/virtio-balloon.c
+++ b/hw/virtio/virtio-balloon.c
@@ -958,7 +958,7 @@ static void virtio_balloon_device_reset(VirtIODevice *vdev)
s->poison_val = 0;
}
-static void virtio_balloon_set_status(VirtIODevice *vdev, uint8_t status)
+static int virtio_balloon_set_status(VirtIODevice *vdev, uint8_t status)
{
VirtIOBalloon *s = VIRTIO_BALLOON(vdev);
@@ -988,6 +988,7 @@ static void virtio_balloon_set_status(VirtIODevice *vdev, uint8_t status)
qemu_mutex_unlock(&s->free_page_lock);
}
}
+ return 0;
}
static ResettableState *virtio_balloon_get_reset_state(Object *obj)
diff --git a/hw/virtio/virtio-crypto.c b/hw/virtio/virtio-crypto.c
index e24d6914b6..517f2089c5 100644
--- a/hw/virtio/virtio-crypto.c
+++ b/hw/virtio/virtio-crypto.c
@@ -1197,11 +1197,12 @@ static void virtio_crypto_vhost_status(VirtIOCrypto *c, uint8_t status)
}
}
-static void virtio_crypto_set_status(VirtIODevice *vdev, uint8_t status)
+static int virtio_crypto_set_status(VirtIODevice *vdev, uint8_t status)
{
VirtIOCrypto *vcrypto = VIRTIO_CRYPTO(vdev);
virtio_crypto_vhost_status(vcrypto, status);
+ return 0;
}
static void virtio_crypto_guest_notifier_mask(VirtIODevice *vdev, int idx,
diff --git a/hw/virtio/virtio-iommu.c b/hw/virtio/virtio-iommu.c
index 54060988ef..3500f1b082 100644
--- a/hw/virtio/virtio-iommu.c
+++ b/hw/virtio/virtio-iommu.c
@@ -1522,9 +1522,10 @@ static void virtio_iommu_device_reset_exit(Object *obj, ResetType type)
NULL, NULL, virtio_iommu_put_endpoint);
}
-static void virtio_iommu_set_status(VirtIODevice *vdev, uint8_t status)
+static int virtio_iommu_set_status(VirtIODevice *vdev, uint8_t status)
{
trace_virtio_iommu_device_status(status);
+ return 0;
}
static void virtio_iommu_instance_init(Object *obj)
diff --git a/hw/virtio/virtio-rng.c b/hw/virtio/virtio-rng.c
index dcb3c71d6a..3df5d2576e 100644
--- a/hw/virtio/virtio-rng.c
+++ b/hw/virtio/virtio-rng.c
@@ -159,17 +159,18 @@ static void check_rate_limit(void *opaque)
vrng->activate_timer = true;
}
-static void virtio_rng_set_status(VirtIODevice *vdev, uint8_t status)
+static int virtio_rng_set_status(VirtIODevice *vdev, uint8_t status)
{
VirtIORNG *vrng = VIRTIO_RNG(vdev);
if (!vdev->vm_running) {
- return;
+ return 0;
}
vdev->status = status;
/* Something changed, try to process buffers */
virtio_rng_process(vrng);
+ return 0;
}
static void virtio_rng_device_realize(DeviceState *dev, Error **errp)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index e3b62522ec..1b55d8d8a2 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -2221,12 +2221,12 @@ int virtio_set_status(VirtIODevice *vdev, uint8_t val)
{
VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
trace_virtio_set_status(vdev, val);
+ int ret = 0;
if (virtio_vdev_has_feature(vdev, VIRTIO_F_VERSION_1)) {
if (!(vdev->status & VIRTIO_CONFIG_S_FEATURES_OK) &&
val & VIRTIO_CONFIG_S_FEATURES_OK) {
- int ret = virtio_validate_features(vdev);
-
+ ret = virtio_validate_features(vdev);
if (ret) {
return ret;
}
@@ -2239,11 +2239,15 @@ int virtio_set_status(VirtIODevice *vdev, uint8_t val)
}
if (k->set_status) {
- k->set_status(vdev, val);
+ ret = k->set_status(vdev, val);
+ if (ret) {
+ qemu_log("set %s status to %d failed, old status: %d\n",
+ vdev->name, val, vdev->status);
+ }
}
vdev->status = val;
- return 0;
+ return ret;
}
static enum virtio_device_endian virtio_default_endian(void)
@@ -3419,7 +3423,7 @@ void virtio_cleanup(VirtIODevice *vdev)
qemu_del_vm_change_state_handler(vdev->vmstate);
}
-static void virtio_vmstate_change(void *opaque, bool running, RunState state)
+static int virtio_vmstate_change(void *opaque, bool running, RunState state)
{
VirtIODevice *vdev = opaque;
BusState *qbus = qdev_get_parent_bus(DEVICE(vdev));
@@ -3436,8 +3440,12 @@ static void virtio_vmstate_change(void *opaque, bool running, RunState state)
}
if (!backend_run) {
- virtio_set_status(vdev, vdev->status);
+ int ret = virtio_set_status(vdev, vdev->status);
+ if (ret) {
+ return ret;
+ }
}
+ return 0;
}
void virtio_instance_init_common(Object *proxy_obj, void *data,
@@ -3489,7 +3497,7 @@ void virtio_init(VirtIODevice *vdev, uint16_t device_id, size_t config_size)
vdev->config = NULL;
}
vdev->vmstate = qdev_add_vm_change_state_handler(DEVICE(vdev),
- virtio_vmstate_change, NULL, vdev);
+ NULL, virtio_vmstate_change, vdev);
vdev->device_endian = virtio_default_endian();
vdev->use_guest_notifier_mask = true;
}
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 21/27] vhost-scsi: support VIRTIO_SCSI_F_HOTPLUG
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (19 preceding siblings ...)
2025-05-14 11:51 ` [PULL 20/27] vhost-user: return failure if backend crash when live migration Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 22/27] virtio: Call set_features during reset Michael S. Tsirkin
` (6 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Dongli Zhang, Stefano Garzarella, Paolo Bonzini,
Fam Zheng
From: Dongli Zhang <dongli.zhang@oracle.com>
So far there isn't way to test host kernel vhost-scsi event queue path,
because VIRTIO_SCSI_F_HOTPLUG isn't supported by QEMU.
virtio-scsi.c and vhost-user-scsi.c already support VIRTIO_SCSI_F_HOTPLUG
as property "hotplug".
Add support to vhost-scsi.c to help evaluate and test event queue.
To test the feature:
1. Create vhost-scsi target with targetcli.
targetcli /backstores/fileio create name=storage file_or_dev=disk01.raw
targetcli /vhost create naa.1123451234512345
targetcli /vhost/naa.1123451234512345/tpg1/luns create /backstores/fileio/storage
2. Create QEMU instance with vhost-scsi.
-device vhost-scsi-pci,wwpn=naa.1123451234512345,hotplug=true
3. Once guest bootup, hotplug a new LUN from host.
targetcli /backstores/fileio create name=storage02 file_or_dev=disk02.raw
targetcli /vhost/naa.1123451234512345/tpg1/luns create /backstores/fileio/storage02
Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
Message-Id: <20250203005215.1502-1-dongli.zhang@oracle.com>
Acked-by: Stefano Garzarella <sgarzare@redhat.com>
---
hw/scsi/vhost-scsi.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/hw/scsi/vhost-scsi.c b/hw/scsi/vhost-scsi.c
index dd4250ebe8..cdf405b0f8 100644
--- a/hw/scsi/vhost-scsi.c
+++ b/hw/scsi/vhost-scsi.c
@@ -359,6 +359,9 @@ static const Property vhost_scsi_properties[] = {
DEFINE_PROP_BIT64("t10_pi", VHostSCSICommon, host_features,
VIRTIO_SCSI_F_T10_PI,
false),
+ DEFINE_PROP_BIT64("hotplug", VHostSCSICommon, host_features,
+ VIRTIO_SCSI_F_HOTPLUG,
+ false),
DEFINE_PROP_BOOL("migratable", VHostSCSICommon, migratable, false),
DEFINE_PROP_BOOL("worker_per_virtqueue", VirtIOSCSICommon,
conf.worker_per_virtqueue, false),
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 22/27] virtio: Call set_features during reset
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (20 preceding siblings ...)
2025-05-14 11:51 ` [PULL 21/27] vhost-scsi: support VIRTIO_SCSI_F_HOTPLUG Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 23/27] virtio: Move virtio_reset() Michael S. Tsirkin
` (5 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Akihiko Odaki, qemu-stable
From: Akihiko Odaki <akihiko.odaki@daynix.com>
virtio-net expects set_features() will be called when the feature set
used by the guest changes to update the number of virtqueues but it is
not called during reset, which will clear all features, leaving the
queues added for VIRTIO_NET_F_MQ or VIRTIO_NET_F_RSS. Not only these
extra queues are visible to the guest, they will cause segmentation
fault during migration.
Call set_features() during reset to remove those queues for virtio-net
as we call set_status(). It will also prevent similar bugs for
virtio-net and other devices in the future.
Fixes: f9d6dbf0bf6e ("virtio-net: remove virtio queues if the guest doesn't support multiqueue")
Buglink: https://issues.redhat.com/browse/RHEL-73842
Cc: qemu-stable@nongnu.org
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250421-reset-v2-1-e4c1ead88ea1@daynix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/virtio/virtio.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 1b55d8d8a2..3300f4afb0 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -2320,6 +2320,8 @@ void virtio_queue_enable(VirtIODevice *vdev, uint32_t queue_index)
}
}
+static int virtio_set_features_nocheck(VirtIODevice *vdev, uint64_t val);
+
void virtio_reset(void *opaque)
{
VirtIODevice *vdev = opaque;
@@ -2350,7 +2352,7 @@ void virtio_reset(void *opaque)
vdev->start_on_kick = false;
vdev->started = false;
vdev->broken = false;
- vdev->guest_features = 0;
+ virtio_set_features_nocheck(vdev, 0);
vdev->queue_sel = 0;
vdev->status = 0;
vdev->disabled = false;
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 23/27] virtio: Move virtio_reset()
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (21 preceding siblings ...)
2025-05-14 11:51 ` [PULL 22/27] virtio: Call set_features during reset Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 24/27] intel_iommu: Use BQL_LOCK_GUARD to manage cleanup automatically Michael S. Tsirkin
` (4 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel; +Cc: Peter Maydell, Akihiko Odaki
From: Akihiko Odaki <akihiko.odaki@daynix.com>
Move virtio_reset() to a later part of the file to remove the forward
declaration of virtio_set_features_nocheck() and to prepare the
situation that we need more operations to perform during reset.
Signed-off-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-Id: <20250421-reset-v2-2-e4c1ead88ea1@daynix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/virtio/virtio.c | 88 ++++++++++++++++++++++------------------------
1 file changed, 43 insertions(+), 45 deletions(-)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index 3300f4afb0..2e98cecf64 100644
--- a/hw/virtio/virtio.c
+++ b/hw/virtio/virtio.c
@@ -2320,51 +2320,6 @@ void virtio_queue_enable(VirtIODevice *vdev, uint32_t queue_index)
}
}
-static int virtio_set_features_nocheck(VirtIODevice *vdev, uint64_t val);
-
-void virtio_reset(void *opaque)
-{
- VirtIODevice *vdev = opaque;
- VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
- int i;
-
- virtio_set_status(vdev, 0);
- if (current_cpu) {
- /* Guest initiated reset */
- vdev->device_endian = virtio_current_cpu_endian();
- } else {
- /* System reset */
- vdev->device_endian = virtio_default_endian();
- }
-
- if (k->get_vhost) {
- struct vhost_dev *hdev = k->get_vhost(vdev);
- /* Only reset when vhost back-end is connected */
- if (hdev && hdev->vhost_ops) {
- vhost_reset_device(hdev);
- }
- }
-
- if (k->reset) {
- k->reset(vdev);
- }
-
- vdev->start_on_kick = false;
- vdev->started = false;
- vdev->broken = false;
- virtio_set_features_nocheck(vdev, 0);
- vdev->queue_sel = 0;
- vdev->status = 0;
- vdev->disabled = false;
- qatomic_set(&vdev->isr, 0);
- vdev->config_vector = VIRTIO_NO_VECTOR;
- virtio_notify_vector(vdev, vdev->config_vector);
-
- for(i = 0; i < VIRTIO_QUEUE_MAX; i++) {
- __virtio_queue_reset(vdev, i);
- }
-}
-
void virtio_queue_set_addr(VirtIODevice *vdev, int n, hwaddr addr)
{
if (!vdev->vq[n].vring.num) {
@@ -3175,6 +3130,49 @@ int virtio_set_features(VirtIODevice *vdev, uint64_t val)
return ret;
}
+void virtio_reset(void *opaque)
+{
+ VirtIODevice *vdev = opaque;
+ VirtioDeviceClass *k = VIRTIO_DEVICE_GET_CLASS(vdev);
+ int i;
+
+ virtio_set_status(vdev, 0);
+ if (current_cpu) {
+ /* Guest initiated reset */
+ vdev->device_endian = virtio_current_cpu_endian();
+ } else {
+ /* System reset */
+ vdev->device_endian = virtio_default_endian();
+ }
+
+ if (k->get_vhost) {
+ struct vhost_dev *hdev = k->get_vhost(vdev);
+ /* Only reset when vhost back-end is connected */
+ if (hdev && hdev->vhost_ops) {
+ vhost_reset_device(hdev);
+ }
+ }
+
+ if (k->reset) {
+ k->reset(vdev);
+ }
+
+ vdev->start_on_kick = false;
+ vdev->started = false;
+ vdev->broken = false;
+ virtio_set_features_nocheck(vdev, 0);
+ vdev->queue_sel = 0;
+ vdev->status = 0;
+ vdev->disabled = false;
+ qatomic_set(&vdev->isr, 0);
+ vdev->config_vector = VIRTIO_NO_VECTOR;
+ virtio_notify_vector(vdev, vdev->config_vector);
+
+ for (i = 0; i < VIRTIO_QUEUE_MAX; i++) {
+ __virtio_queue_reset(vdev, i);
+ }
+}
+
static void virtio_device_check_notification_compatibility(VirtIODevice *vdev,
Error **errp)
{
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 24/27] intel_iommu: Use BQL_LOCK_GUARD to manage cleanup automatically
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (22 preceding siblings ...)
2025-05-14 11:51 ` [PULL 23/27] virtio: Move virtio_reset() Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 25/27] intel_iommu: Take locks when looking for and creating address spaces Michael S. Tsirkin
` (3 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, CLEMENT MATHIEU--DRIF, Zhenzhong Duan, Jason Wang,
Yi Liu, Marcel Apfelbaum, Paolo Bonzini, Richard Henderson,
Eduardo Habkost
From: CLEMENT MATHIEU--DRIF <clement.mathieu--drif@eviden.com>
vtd_switch_address_space needs to take the BQL if not already held.
Use BQL_LOCK_GUARD to make the iommu implementation more consistent.
Signed-off-by: Clement Mathieu--Drif <clement.mathieu--drif@eviden.com>
Message-Id: <20250430124750.240412-2-clement.mathieu--drif@eviden.com>
Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/i386/intel_iommu.c | 10 +---------
1 file changed, 1 insertion(+), 9 deletions(-)
diff --git a/hw/i386/intel_iommu.c b/hw/i386/intel_iommu.c
index 5f8ed1243d..b925e65b02 100644
--- a/hw/i386/intel_iommu.c
+++ b/hw/i386/intel_iommu.c
@@ -1728,8 +1728,6 @@ static bool vtd_as_pt_enabled(VTDAddressSpace *as)
static bool vtd_switch_address_space(VTDAddressSpace *as)
{
bool use_iommu, pt;
- /* Whether we need to take the BQL on our own */
- bool take_bql = !bql_locked();
assert(as);
@@ -1746,9 +1744,7 @@ static bool vtd_switch_address_space(VTDAddressSpace *as)
* from vtd_pt_enable_fast_path(). However the memory APIs need
* it. We'd better make sure we have had it already, or, take it.
*/
- if (take_bql) {
- bql_lock();
- }
+ BQL_LOCK_GUARD();
/* Turn off first then on the other */
if (use_iommu) {
@@ -1801,10 +1797,6 @@ static bool vtd_switch_address_space(VTDAddressSpace *as)
memory_region_set_enabled(&as->iommu_ir_fault, false);
}
- if (take_bql) {
- bql_unlock();
- }
-
return use_iommu;
}
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 25/27] intel_iommu: Take locks when looking for and creating address spaces
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (23 preceding siblings ...)
2025-05-14 11:51 ` [PULL 24/27] intel_iommu: Use BQL_LOCK_GUARD to manage cleanup automatically Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 26/27] hw/i386/amd_iommu: Isolate AMDVI-PCI from amd-iommu device to allow full control over the PCI device creation Michael S. Tsirkin
` (2 subsequent siblings)
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, CLEMENT MATHIEU--DRIF, Zhenzhong Duan, Jason Wang,
Yi Liu, Marcel Apfelbaum, Paolo Bonzini, Richard Henderson,
Eduardo Habkost
From: CLEMENT MATHIEU--DRIF <clement.mathieu--drif@eviden.com>
vtd_find_add_as can be called by multiple threads which leads to a race
condition. Taking the IOMMU lock ensures we avoid such a race.
Moreover we also need to take the bql to avoid an assert to fail in
memory_region_add_subregion_overlap when actually allocating a new
address space.
Signed-off-by: Clement Mathieu--Drif <clement.mathieu--drif@eviden.com>
Message-Id: <20250430124750.240412-3-clement.mathieu--drif@eviden.com>
Reviewed-by: Zhenzhong Duan <zhenzhong.duan@intel.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/i386/intel_iommu.c | 25 ++++++++++++++++++++++++-
1 file changed, 24 insertions(+), 1 deletion(-)
diff --git a/hw/i386/intel_iommu.c b/hw/i386/intel_iommu.c
index b925e65b02..69d72ad35c 100644
--- a/hw/i386/intel_iommu.c
+++ b/hw/i386/intel_iommu.c
@@ -4205,9 +4205,30 @@ VTDAddressSpace *vtd_find_add_as(IntelIOMMUState *s, PCIBus *bus,
VTDAddressSpace *vtd_dev_as;
char name[128];
+ vtd_iommu_lock(s);
vtd_dev_as = g_hash_table_lookup(s->vtd_address_spaces, &key);
+ vtd_iommu_unlock(s);
+
if (!vtd_dev_as) {
- struct vtd_as_key *new_key = g_malloc(sizeof(*new_key));
+ struct vtd_as_key *new_key;
+ /* Slow path */
+
+ /*
+ * memory_region_add_subregion_overlap requires the bql,
+ * make sure we own it.
+ */
+ BQL_LOCK_GUARD();
+ vtd_iommu_lock(s);
+
+ /* Check again as we released the lock for a moment */
+ vtd_dev_as = g_hash_table_lookup(s->vtd_address_spaces, &key);
+ if (vtd_dev_as) {
+ vtd_iommu_unlock(s);
+ return vtd_dev_as;
+ }
+
+ /* Still nothing, allocate a new address space */
+ new_key = g_malloc(sizeof(*new_key));
new_key->bus = bus;
new_key->devfn = devfn;
@@ -4298,6 +4319,8 @@ VTDAddressSpace *vtd_find_add_as(IntelIOMMUState *s, PCIBus *bus,
vtd_switch_address_space(vtd_dev_as);
g_hash_table_insert(s->vtd_address_spaces, new_key, vtd_dev_as);
+
+ vtd_iommu_unlock(s);
}
return vtd_dev_as;
}
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 26/27] hw/i386/amd_iommu: Isolate AMDVI-PCI from amd-iommu device to allow full control over the PCI device creation
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (24 preceding siblings ...)
2025-05-14 11:51 ` [PULL 25/27] intel_iommu: Take locks when looking for and creating address spaces Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-14 11:51 ` [PULL 27/27] hw/i386/amd_iommu: Allow migration when explicitly create the AMDVI-PCI device Michael S. Tsirkin
2025-05-15 21:53 ` [PULL 00/27] virtio,pci,pc: fixes, features Stefan Hajnoczi
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Suravee Suthikulpanit, Daniel P . Berrangé,
Igor Mammedov, Ani Sinha, Paolo Bonzini, Richard Henderson,
Eduardo Habkost, Marcel Apfelbaum
From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Current amd-iommu model internally creates an AMDVI-PCI device. Here is
a snippet from info qtree:
bus: main-system-bus
type System
dev: amd-iommu, id ""
xtsup = false
pci-id = ""
intremap = "on"
device-iotlb = false
pt = true
...
dev: q35-pcihost, id ""
MCFG = -1 (0xffffffffffffffff)
pci-hole64-size = 34359738368 (32 GiB)
below-4g-mem-size = 134217728 (128 MiB)
above-4g-mem-size = 0 (0 B)
smm-ranges = true
x-pci-hole64-fix = true
x-config-reg-migration-enabled = true
bypass-iommu = false
bus: pcie.0
type PCIE
dev: AMDVI-PCI, id ""
addr = 01.0
romfile = ""
romsize = 4294967295 (0xffffffff)
rombar = -1 (0xffffffffffffffff)
multifunction = false
x-pcie-lnksta-dllla = true
x-pcie-extcap-init = true
failover_pair_id = ""
acpi-index = 0 (0x0)
x-pcie-err-unc-mask = true
x-pcie-ari-nextfn-1 = false
x-max-bounce-buffer-size = 4096 (4 KiB)
x-pcie-ext-tag = true
busnr = 0 (0x0)
class Class 0806, addr 00:01.0, pci id 1022:0000 (sub 1af4:1100)
...
This prohibits users from specifying the PCI topology for the amd-iommu device,
which becomes a problem when trying to support VM migration since it does not
guarantee the same enumeration of AMD IOMMU device.
Therefore, allow the 'AMDVI-PCI' device to optionally be pre-created and
associated with a 'amd-iommu' device via a new 'pci-id' parameter on the
latter.
For example:
-device AMDVI-PCI,id=iommupci0,bus=pcie.0,addr=0x05 \
-device amd-iommu,intremap=on,pt=on,xtsup=on,pci-id=iommupci0 \
For backward-compatibility, internally create the AMDVI-PCI device if not
specified on the CLI.
Co-developed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Message-Id: <20250504170405.12623-2-suravee.suthikulpanit@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/i386/amd_iommu.h | 3 ++-
hw/i386/acpi-build.c | 8 +++----
hw/i386/amd_iommu.c | 53 ++++++++++++++++++++++++++------------------
3 files changed, 38 insertions(+), 26 deletions(-)
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index 28125130c6..7a28181d9c 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -315,7 +315,8 @@ struct AMDVIPCIState {
struct AMDVIState {
X86IOMMUState iommu; /* IOMMU bus device */
- AMDVIPCIState pci; /* IOMMU PCI device */
+ AMDVIPCIState *pci; /* IOMMU PCI device */
+ char *pci_id; /* ID of AMDVI-PCI device, if user created */
uint32_t version;
diff --git a/hw/i386/acpi-build.c b/hw/i386/acpi-build.c
index f40ad062f9..61851cc840 100644
--- a/hw/i386/acpi-build.c
+++ b/hw/i386/acpi-build.c
@@ -2333,10 +2333,10 @@ build_amd_iommu(GArray *table_data, BIOSLinker *linker, const char *oem_id,
build_append_int_noprefix(table_data, ivhd_blob->len + 24, 2);
/* DeviceID */
build_append_int_noprefix(table_data,
- object_property_get_int(OBJECT(&s->pci), "addr",
+ object_property_get_int(OBJECT(s->pci), "addr",
&error_abort), 2);
/* Capability offset */
- build_append_int_noprefix(table_data, s->pci.capab_offset, 2);
+ build_append_int_noprefix(table_data, s->pci->capab_offset, 2);
/* IOMMU base address */
build_append_int_noprefix(table_data, s->mr_mmio.addr, 8);
/* PCI Segment Group */
@@ -2368,10 +2368,10 @@ build_amd_iommu(GArray *table_data, BIOSLinker *linker, const char *oem_id,
build_append_int_noprefix(table_data, ivhd_blob->len + 40, 2);
/* DeviceID */
build_append_int_noprefix(table_data,
- object_property_get_int(OBJECT(&s->pci), "addr",
+ object_property_get_int(OBJECT(s->pci), "addr",
&error_abort), 2);
/* Capability offset */
- build_append_int_noprefix(table_data, s->pci.capab_offset, 2);
+ build_append_int_noprefix(table_data, s->pci->capab_offset, 2);
/* IOMMU base address */
build_append_int_noprefix(table_data, s->mr_mmio.addr, 8);
/* PCI Segment Group */
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index 2cf7e24a21..f5466fdc98 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -167,11 +167,11 @@ static void amdvi_generate_msi_interrupt(AMDVIState *s)
{
MSIMessage msg = {};
MemTxAttrs attrs = {
- .requester_id = pci_requester_id(&s->pci.dev)
+ .requester_id = pci_requester_id(&s->pci->dev)
};
- if (msi_enabled(&s->pci.dev)) {
- msg = msi_get_message(&s->pci.dev, 0);
+ if (msi_enabled(&s->pci->dev)) {
+ msg = msi_get_message(&s->pci->dev, 0);
address_space_stl_le(&address_space_memory, msg.address, msg.data,
attrs, NULL);
}
@@ -239,7 +239,7 @@ static void amdvi_page_fault(AMDVIState *s, uint16_t devid,
info |= AMDVI_EVENT_IOPF_I | AMDVI_EVENT_IOPF;
amdvi_encode_event(evt, devid, addr, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
/*
@@ -256,7 +256,7 @@ static void amdvi_log_devtab_error(AMDVIState *s, uint16_t devid,
amdvi_encode_event(evt, devid, devtab, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
/* log an event trying to access command buffer
@@ -269,7 +269,7 @@ static void amdvi_log_command_error(AMDVIState *s, hwaddr addr)
amdvi_encode_event(evt, 0, addr, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
/* log an illegal command event
@@ -310,7 +310,7 @@ static void amdvi_log_pagetab_error(AMDVIState *s, uint16_t devid,
info |= AMDVI_EVENT_PAGE_TAB_HW_ERROR;
amdvi_encode_event(evt, devid, addr, info);
amdvi_log_event(s, evt);
- pci_word_test_and_set_mask(s->pci.dev.config + PCI_STATUS,
+ pci_word_test_and_set_mask(s->pci->dev.config + PCI_STATUS,
PCI_STATUS_SIG_TARGET_ABORT);
}
@@ -1607,7 +1607,7 @@ static void amdvi_sysbus_reset(DeviceState *dev)
{
AMDVIState *s = AMD_IOMMU_DEVICE(dev);
- msi_reset(&s->pci.dev);
+ msi_reset(&s->pci->dev);
amdvi_init(s);
}
@@ -1619,14 +1619,32 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
X86MachineState *x86ms = X86_MACHINE(ms);
PCIBus *bus = pcms->pcibus;
+ if (s->pci_id) {
+ PCIDevice *pdev = NULL;
+ int ret = pci_qdev_find_device(s->pci_id, &pdev);
+
+ if (ret) {
+ error_report("Cannot find PCI device '%s'", s->pci_id);
+ return;
+ }
+
+ if (!object_dynamic_cast(OBJECT(pdev), TYPE_AMD_IOMMU_PCI)) {
+ error_report("Device '%s' must be an AMDVI-PCI device type", s->pci_id);
+ return;
+ }
+
+ s->pci = AMD_IOMMU_PCI(pdev);
+ } else {
+ s->pci = AMD_IOMMU_PCI(object_new(TYPE_AMD_IOMMU_PCI));
+ /* This device should take care of IOMMU PCI properties */
+ if (!qdev_realize(DEVICE(s->pci), &bus->qbus, errp)) {
+ return;
+ }
+ }
+
s->iotlb = g_hash_table_new_full(amdvi_uint64_hash,
amdvi_uint64_equal, g_free, g_free);
- /* This device should take care of IOMMU PCI properties */
- if (!qdev_realize(DEVICE(&s->pci), &bus->qbus, errp)) {
- return;
- }
-
/* Pseudo address space under root PCI bus. */
x86ms->ioapic_as = amdvi_host_dma_iommu(bus, s, AMDVI_IOAPIC_SB_DEVID);
@@ -1663,6 +1681,7 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
static const Property amdvi_properties[] = {
DEFINE_PROP_BOOL("xtsup", AMDVIState, xtsup, false),
+ DEFINE_PROP_STRING("pci-id", AMDVIState, pci_id),
};
static const VMStateDescription vmstate_amdvi_sysbus = {
@@ -1670,13 +1689,6 @@ static const VMStateDescription vmstate_amdvi_sysbus = {
.unmigratable = 1
};
-static void amdvi_sysbus_instance_init(Object *klass)
-{
- AMDVIState *s = AMD_IOMMU_DEVICE(klass);
-
- object_initialize(&s->pci, sizeof(s->pci), TYPE_AMD_IOMMU_PCI);
-}
-
static void amdvi_sysbus_class_init(ObjectClass *klass, const void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
@@ -1696,7 +1708,6 @@ static const TypeInfo amdvi_sysbus = {
.name = TYPE_AMD_IOMMU_DEVICE,
.parent = TYPE_X86_IOMMU_DEVICE,
.instance_size = sizeof(AMDVIState),
- .instance_init = amdvi_sysbus_instance_init,
.class_init = amdvi_sysbus_class_init
};
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* [PULL 27/27] hw/i386/amd_iommu: Allow migration when explicitly create the AMDVI-PCI device
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (25 preceding siblings ...)
2025-05-14 11:51 ` [PULL 26/27] hw/i386/amd_iommu: Isolate AMDVI-PCI from amd-iommu device to allow full control over the PCI device creation Michael S. Tsirkin
@ 2025-05-14 11:51 ` Michael S. Tsirkin
2025-05-15 21:53 ` [PULL 00/27] virtio,pci,pc: fixes, features Stefan Hajnoczi
27 siblings, 0 replies; 31+ messages in thread
From: Michael S. Tsirkin @ 2025-05-14 11:51 UTC (permalink / raw)
To: qemu-devel
Cc: Peter Maydell, Suravee Suthikulpanit, Marcel Apfelbaum,
Paolo Bonzini, Richard Henderson, Eduardo Habkost
From: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Add migration support for AMD IOMMU model by saving necessary AMDVIState
parameters for MMIO registers, device table, command buffer, and event
buffers.
Also change devtab_len type from size_t to uint64_t to avoid 32-bit build
issue.
Signed-off-by: Suravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Message-Id: <20250504170405.12623-3-suravee.suthikulpanit@amd.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
---
hw/i386/amd_iommu.h | 2 +-
hw/i386/amd_iommu.c | 48 +++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 49 insertions(+), 1 deletion(-)
diff --git a/hw/i386/amd_iommu.h b/hw/i386/amd_iommu.h
index 7a28181d9c..5672bdef89 100644
--- a/hw/i386/amd_iommu.h
+++ b/hw/i386/amd_iommu.h
@@ -329,7 +329,7 @@ struct AMDVIState {
bool excl_enabled;
hwaddr devtab; /* base address device table */
- size_t devtab_len; /* device table length */
+ uint64_t devtab_len; /* device table length */
hwaddr cmdbuf; /* command buffer base address */
uint64_t cmdbuf_len; /* command buffer length */
diff --git a/hw/i386/amd_iommu.c b/hw/i386/amd_iommu.c
index f5466fdc98..0775c8f3bb 100644
--- a/hw/i386/amd_iommu.c
+++ b/hw/i386/amd_iommu.c
@@ -1611,8 +1611,55 @@ static void amdvi_sysbus_reset(DeviceState *dev)
amdvi_init(s);
}
+static const VMStateDescription vmstate_amdvi_sysbus_migratable = {
+ .name = "amd-iommu",
+ .version_id = 1,
+ .minimum_version_id = 1,
+ .priority = MIG_PRI_IOMMU,
+ .fields = (VMStateField[]) {
+ /* Updated in amdvi_handle_control_write() */
+ VMSTATE_BOOL(enabled, AMDVIState),
+ VMSTATE_BOOL(ga_enabled, AMDVIState),
+ VMSTATE_BOOL(ats_enabled, AMDVIState),
+ VMSTATE_BOOL(cmdbuf_enabled, AMDVIState),
+ VMSTATE_BOOL(completion_wait_intr, AMDVIState),
+ VMSTATE_BOOL(evtlog_enabled, AMDVIState),
+ VMSTATE_BOOL(evtlog_intr, AMDVIState),
+ /* Updated in amdvi_handle_devtab_write() */
+ VMSTATE_UINT64(devtab, AMDVIState),
+ VMSTATE_UINT64(devtab_len, AMDVIState),
+ /* Updated in amdvi_handle_cmdbase_write() */
+ VMSTATE_UINT64(cmdbuf, AMDVIState),
+ VMSTATE_UINT64(cmdbuf_len, AMDVIState),
+ /* Updated in amdvi_handle_cmdhead_write() */
+ VMSTATE_UINT32(cmdbuf_head, AMDVIState),
+ /* Updated in amdvi_handle_cmdtail_write() */
+ VMSTATE_UINT32(cmdbuf_tail, AMDVIState),
+ /* Updated in amdvi_handle_evtbase_write() */
+ VMSTATE_UINT64(evtlog, AMDVIState),
+ VMSTATE_UINT32(evtlog_len, AMDVIState),
+ /* Updated in amdvi_handle_evthead_write() */
+ VMSTATE_UINT32(evtlog_head, AMDVIState),
+ /* Updated in amdvi_handle_evttail_write() */
+ VMSTATE_UINT32(evtlog_tail, AMDVIState),
+ /* Updated in amdvi_handle_pprbase_write() */
+ VMSTATE_UINT64(ppr_log, AMDVIState),
+ VMSTATE_UINT32(pprlog_len, AMDVIState),
+ /* Updated in amdvi_handle_pprhead_write() */
+ VMSTATE_UINT32(pprlog_head, AMDVIState),
+ /* Updated in amdvi_handle_tailhead_write() */
+ VMSTATE_UINT32(pprlog_tail, AMDVIState),
+ /* MMIO registers */
+ VMSTATE_UINT8_ARRAY(mmior, AMDVIState, AMDVI_MMIO_SIZE),
+ VMSTATE_UINT8_ARRAY(romask, AMDVIState, AMDVI_MMIO_SIZE),
+ VMSTATE_UINT8_ARRAY(w1cmask, AMDVIState, AMDVI_MMIO_SIZE),
+ VMSTATE_END_OF_LIST()
+ }
+};
+
static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
{
+ DeviceClass *dc = (DeviceClass *) object_get_class(OBJECT(dev));
AMDVIState *s = AMD_IOMMU_DEVICE(dev);
MachineState *ms = MACHINE(qdev_get_machine());
PCMachineState *pcms = PC_MACHINE(ms);
@@ -1634,6 +1681,7 @@ static void amdvi_sysbus_realize(DeviceState *dev, Error **errp)
}
s->pci = AMD_IOMMU_PCI(pdev);
+ dc->vmsd = &vmstate_amdvi_sysbus_migratable;
} else {
s->pci = AMD_IOMMU_PCI(object_new(TYPE_AMD_IOMMU_PCI));
/* This device should take care of IOMMU PCI properties */
--
MST
^ permalink raw reply related [flat|nested] 31+ messages in thread* Re: [PULL 00/27] virtio,pci,pc: fixes, features
2025-05-14 11:50 [PULL 00/27] virtio,pci,pc: fixes, features Michael S. Tsirkin
` (26 preceding siblings ...)
2025-05-14 11:51 ` [PULL 27/27] hw/i386/amd_iommu: Allow migration when explicitly create the AMDVI-PCI device Michael S. Tsirkin
@ 2025-05-15 21:53 ` Stefan Hajnoczi
27 siblings, 0 replies; 31+ messages in thread
From: Stefan Hajnoczi @ 2025-05-15 21:53 UTC (permalink / raw)
To: Michael S. Tsirkin; +Cc: qemu-devel, Peter Maydell
[-- Attachment #1: Type: text/plain, Size: 116 bytes --]
Applied, thanks.
Please update the changelog at https://wiki.qemu.org/ChangeLog/10.1 for any user-visible changes.
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]
^ permalink raw reply [flat|nested] 31+ messages in thread