From 9547754f40ee5c5e3d1dbed0fbc972caacd075e8 Mon Sep 17 00:00:00 2001 From: Jonathan Cameron Date: Fri, 26 May 2023 18:00:08 +0100 Subject: hw/cxl: QMP based poison injection support Inject poison using QMP command cxl-inject-poison to add an entry to the poison list. For now, the poison is not returned CXL.mem reads, but only via the mailbox command Get Poison List. So a normal memory read to an address that is on the poison list will not yet result in a synchronous exception (and similar for partial cacheline writes). That is left for a future patch. See CXL rev 3.0, sec 8.2.9.8.4.1 Get Poison list (Opcode 4300h) Kernel patches to use this interface here: https://lore.kernel.org/linux-cxl/cover.1665606782.git.alison.schofield@intel.com/ To inject poison using QMP (telnet to the QMP port) { "execute": "qmp_capabilities" } { "execute": "cxl-inject-poison", "arguments": { "path": "/machine/peripheral/cxl-pmem0", "start": 2048, "length": 256 } } Adjusted to select a device on your machine. Note that the poison list supported is kept short enough to avoid the complexity of state machine that is needed to handle the MORE flag. Reviewed-by: Fan Ni Reviewed-by: Ira Weiny Acked-by: Markus Armbruster Signed-off-by: Jonathan Cameron Message-Id: <20230526170010.574-3-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/mem/cxl_type3.c | 56 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 56 insertions(+) (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index 2adacbd01b..ab600735eb 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -947,6 +947,62 @@ static void set_lsa(CXLType3Dev *ct3d, const void *buf, uint64_t size, */ } +void cxl_set_poison_list_overflowed(CXLType3Dev *ct3d) +{ + ct3d->poison_list_overflowed = true; + ct3d->poison_list_overflow_ts = + cxl_device_get_timestamp(&ct3d->cxl_dstate); +} + +void qmp_cxl_inject_poison(const char *path, uint64_t start, uint64_t length, + Error **errp) +{ + Object *obj = object_resolve_path(path, NULL); + CXLType3Dev *ct3d; + CXLPoison *p; + + if (length % 64) { + error_setg(errp, "Poison injection must be in multiples of 64 bytes"); + return; + } + if (start % 64) { + error_setg(errp, "Poison start address must be 64 byte aligned"); + return; + } + if (!obj) { + error_setg(errp, "Unable to resolve path"); + return; + } + if (!object_dynamic_cast(obj, TYPE_CXL_TYPE3)) { + error_setg(errp, "Path does not point to a CXL type 3 device"); + return; + } + + ct3d = CXL_TYPE3(obj); + + QLIST_FOREACH(p, &ct3d->poison_list, node) { + if (((start >= p->start) && (start < p->start + p->length)) || + ((start + length > p->start) && + (start + length <= p->start + p->length))) { + error_setg(errp, "Overlap with existing poisoned region not supported"); + return; + } + } + + if (ct3d->poison_list_cnt == CXL_POISON_LIST_LIMIT) { + cxl_set_poison_list_overflowed(ct3d); + return; + } + + p = g_new0(CXLPoison, 1); + p->length = length; + p->start = start; + p->type = CXL_POISON_TYPE_INTERNAL; /* Different from injected via the mbox */ + + QLIST_INSERT_HEAD(&ct3d->poison_list, p, node); + ct3d->poison_list_cnt++; +} + /* For uncorrectable errors include support for multiple header recording */ void qmp_cxl_inject_uncorrectable_errors(const char *path, CXLUncorErrorRecordList *errors, -- cgit 1.4.1 From 6bda41a69bdcee8ff7dcf75df2f9647ce55908ab Mon Sep 17 00:00:00 2001 From: Jonathan Cameron Date: Fri, 26 May 2023 18:00:10 +0100 Subject: hw/cxl: Add clear poison mailbox command support. Current implementation is very simple so many of the corner cases do not exist (e.g. fragmenting larger poison list entries) Reviewed-by: Fan Ni Reviewed-by: Ira Weiny Signed-off-by: Jonathan Cameron Message-Id: <20230526170010.574-5-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/cxl/cxl-mailbox-utils.c | 82 +++++++++++++++++++++++++++++++++++++++++++++ hw/mem/cxl_type3.c | 37 ++++++++++++++++++++ include/hw/cxl/cxl_device.h | 1 + 3 files changed, 120 insertions(+) (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c index 6c476ad7f4..e3401b6be8 100644 --- a/hw/cxl/cxl-mailbox-utils.c +++ b/hw/cxl/cxl-mailbox-utils.c @@ -65,6 +65,7 @@ enum { MEDIA_AND_POISON = 0x43, #define GET_POISON_LIST 0x0 #define INJECT_POISON 0x1 + #define CLEAR_POISON 0x2 }; /* 8.2.8.4.5.1 Command Return Codes */ @@ -512,6 +513,85 @@ static CXLRetCode cmd_media_inject_poison(struct cxl_cmd *cmd, return CXL_MBOX_SUCCESS; } +static CXLRetCode cmd_media_clear_poison(struct cxl_cmd *cmd, + CXLDeviceState *cxl_dstate, + uint16_t *len_unused) +{ + CXLType3Dev *ct3d = container_of(cxl_dstate, CXLType3Dev, cxl_dstate); + CXLPoisonList *poison_list = &ct3d->poison_list; + CXLType3Class *cvc = CXL_TYPE3_GET_CLASS(ct3d); + struct clear_poison_pl { + uint64_t dpa; + uint8_t data[64]; + }; + CXLPoison *ent; + uint64_t dpa; + + struct clear_poison_pl *in = (void *)cmd->payload; + + dpa = ldq_le_p(&in->dpa); + if (dpa + CXL_CACHE_LINE_SIZE > cxl_dstate->mem_size) { + return CXL_MBOX_INVALID_PA; + } + + /* Clearing a region with no poison is not an error so always do so */ + if (cvc->set_cacheline) { + if (!cvc->set_cacheline(ct3d, dpa, in->data)) { + return CXL_MBOX_INTERNAL_ERROR; + } + } + + QLIST_FOREACH(ent, poison_list, node) { + /* + * Test for contained in entry. Simpler than general case + * as clearing 64 bytes and entries 64 byte aligned + */ + if ((dpa >= ent->start) && (dpa < ent->start + ent->length)) { + break; + } + } + if (!ent) { + return CXL_MBOX_SUCCESS; + } + + QLIST_REMOVE(ent, node); + ct3d->poison_list_cnt--; + + if (dpa > ent->start) { + CXLPoison *frag; + /* Cannot overflow as replacing existing entry */ + + frag = g_new0(CXLPoison, 1); + + frag->start = ent->start; + frag->length = dpa - ent->start; + frag->type = ent->type; + + QLIST_INSERT_HEAD(poison_list, frag, node); + ct3d->poison_list_cnt++; + } + + if (dpa + CXL_CACHE_LINE_SIZE < ent->start + ent->length) { + CXLPoison *frag; + + if (ct3d->poison_list_cnt == CXL_POISON_LIST_LIMIT) { + cxl_set_poison_list_overflowed(ct3d); + } else { + frag = g_new0(CXLPoison, 1); + + frag->start = dpa + CXL_CACHE_LINE_SIZE; + frag->length = ent->start + ent->length - frag->start; + frag->type = ent->type; + QLIST_INSERT_HEAD(poison_list, frag, node); + ct3d->poison_list_cnt++; + } + } + /* Any fragments have been added, free original entry */ + g_free(ent); + + return CXL_MBOX_SUCCESS; +} + #define IMMEDIATE_CONFIG_CHANGE (1 << 1) #define IMMEDIATE_DATA_CHANGE (1 << 2) #define IMMEDIATE_POLICY_CHANGE (1 << 3) @@ -543,6 +623,8 @@ static struct cxl_cmd cxl_cmd_set[256][256] = { cmd_media_get_poison_list, 16, 0 }, [MEDIA_AND_POISON][INJECT_POISON] = { "MEDIA_AND_POISON_INJECT_POISON", cmd_media_inject_poison, 8, 0 }, + [MEDIA_AND_POISON][CLEAR_POISON] = { "MEDIA_AND_POISON_CLEAR_POISON", + cmd_media_clear_poison, 72, 0 }, }; void cxl_process_mailbox(CXLDeviceState *cxl_dstate) diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index ab600735eb..d751803188 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -947,6 +947,42 @@ static void set_lsa(CXLType3Dev *ct3d, const void *buf, uint64_t size, */ } +static bool set_cacheline(CXLType3Dev *ct3d, uint64_t dpa_offset, uint8_t *data) +{ + MemoryRegion *vmr = NULL, *pmr = NULL; + AddressSpace *as; + + if (ct3d->hostvmem) { + vmr = host_memory_backend_get_memory(ct3d->hostvmem); + } + if (ct3d->hostpmem) { + pmr = host_memory_backend_get_memory(ct3d->hostpmem); + } + + if (!vmr && !pmr) { + return false; + } + + if (dpa_offset + CXL_CACHE_LINE_SIZE > ct3d->cxl_dstate.mem_size) { + return false; + } + + if (vmr) { + if (dpa_offset < memory_region_size(vmr)) { + as = &ct3d->hostvmem_as; + } else { + as = &ct3d->hostpmem_as; + dpa_offset -= memory_region_size(vmr); + } + } else { + as = &ct3d->hostpmem_as; + } + + address_space_write(as, dpa_offset, MEMTXATTRS_UNSPECIFIED, &data, + CXL_CACHE_LINE_SIZE); + return true; +} + void cxl_set_poison_list_overflowed(CXLType3Dev *ct3d) { ct3d->poison_list_overflowed = true; @@ -1168,6 +1204,7 @@ static void ct3_class_init(ObjectClass *oc, void *data) cvc->get_lsa_size = get_lsa_size; cvc->get_lsa = get_lsa; cvc->set_lsa = set_lsa; + cvc->set_cacheline = set_cacheline; } static const TypeInfo ct3d_info = { diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h index 32c234ea91..73328a52cf 100644 --- a/include/hw/cxl/cxl_device.h +++ b/include/hw/cxl/cxl_device.h @@ -298,6 +298,7 @@ struct CXLType3Class { uint64_t offset); void (*set_lsa)(CXLType3Dev *ct3d, const void *buf, uint64_t size, uint64_t offset); + bool (*set_cacheline)(CXLType3Dev *ct3d, uint64_t dpa_offset, uint8_t *data); }; MemTxResult cxl_type3_read(PCIDevice *d, hwaddr host_addr, uint64_t *data, -- cgit 1.4.1 From 22d7e3be0714f39bae43bd0c05f6e6d149a47b13 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Tue, 30 May 2023 14:35:59 +0100 Subject: hw/cxl/events: Wire up get/clear event mailbox commands CXL testing is benefited from an artificial event log injection mechanism. Add an event log infrastructure to insert, get, and clear events from the various logs available on a device. Replace the stubbed out CXL Get/Clear Event mailbox commands with commands that operate on the new infrastructure. Signed-off-by: Ira Weiny Reviewed-by: Fan Ni Signed-off-by: Jonathan Cameron Message-Id: <20230530133603.16934-4-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/cxl/cxl-events.c | 217 ++++++++++++++++++++++++++++++++++++++++++++ hw/cxl/cxl-mailbox-utils.c | 40 +++++++- hw/cxl/meson.build | 1 + hw/mem/cxl_type3.c | 1 + include/hw/cxl/cxl_device.h | 25 +++++ include/hw/cxl/cxl_events.h | 55 +++++++++++ 6 files changed, 337 insertions(+), 2 deletions(-) create mode 100644 hw/cxl/cxl-events.c (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/cxl/cxl-events.c b/hw/cxl/cxl-events.c new file mode 100644 index 0000000000..5da1b76b97 --- /dev/null +++ b/hw/cxl/cxl-events.c @@ -0,0 +1,217 @@ +/* + * CXL Event processing + * + * Copyright(C) 2023 Intel Corporation. + * + * This work is licensed under the terms of the GNU GPL, version 2. See the + * COPYING file in the top-level directory. + */ + +#include + +#include "qemu/osdep.h" +#include "qemu/bswap.h" +#include "qemu/typedefs.h" +#include "qemu/error-report.h" +#include "hw/cxl/cxl.h" +#include "hw/cxl/cxl_events.h" + +/* Artificial limit on the number of events a log can hold */ +#define CXL_TEST_EVENT_OVERFLOW 8 + +static void reset_overflow(CXLEventLog *log) +{ + log->overflow_err_count = 0; + log->first_overflow_timestamp = 0; + log->last_overflow_timestamp = 0; +} + +void cxl_event_init(CXLDeviceState *cxlds) +{ + CXLEventLog *log; + int i; + + for (i = 0; i < CXL_EVENT_TYPE_MAX; i++) { + log = &cxlds->event_logs[i]; + log->next_handle = 1; + log->overflow_err_count = 0; + log->first_overflow_timestamp = 0; + log->last_overflow_timestamp = 0; + qemu_mutex_init(&log->lock); + QSIMPLEQ_INIT(&log->events); + } +} + +static CXLEvent *cxl_event_get_head(CXLEventLog *log) +{ + return QSIMPLEQ_FIRST(&log->events); +} + +static CXLEvent *cxl_event_get_next(CXLEvent *entry) +{ + return QSIMPLEQ_NEXT(entry, node); +} + +static int cxl_event_count(CXLEventLog *log) +{ + CXLEvent *event; + int rc = 0; + + QSIMPLEQ_FOREACH(event, &log->events, node) { + rc++; + } + + return rc; +} + +static bool cxl_event_empty(CXLEventLog *log) +{ + return QSIMPLEQ_EMPTY(&log->events); +} + +static void cxl_event_delete_head(CXLDeviceState *cxlds, + CXLEventLogType log_type, + CXLEventLog *log) +{ + CXLEvent *entry = cxl_event_get_head(log); + + reset_overflow(log); + QSIMPLEQ_REMOVE_HEAD(&log->events, node); + if (cxl_event_empty(log)) { + cxl_event_set_status(cxlds, log_type, false); + } + g_free(entry); +} + +/* + * return true if an interrupt should be generated as a result + * of inserting this event. + */ +bool cxl_event_insert(CXLDeviceState *cxlds, CXLEventLogType log_type, + CXLEventRecordRaw *event) +{ + uint64_t time; + CXLEventLog *log; + CXLEvent *entry; + + if (log_type >= CXL_EVENT_TYPE_MAX) { + return false; + } + + time = cxl_device_get_timestamp(cxlds); + + log = &cxlds->event_logs[log_type]; + + QEMU_LOCK_GUARD(&log->lock); + + if (cxl_event_count(log) >= CXL_TEST_EVENT_OVERFLOW) { + if (log->overflow_err_count == 0) { + log->first_overflow_timestamp = time; + } + log->overflow_err_count++; + log->last_overflow_timestamp = time; + return false; + } + + entry = g_new0(CXLEvent, 1); + + memcpy(&entry->data, event, sizeof(*event)); + + entry->data.hdr.handle = cpu_to_le16(log->next_handle); + log->next_handle++; + /* 0 handle is never valid */ + if (log->next_handle == 0) { + log->next_handle++; + } + entry->data.hdr.timestamp = cpu_to_le64(time); + + QSIMPLEQ_INSERT_TAIL(&log->events, entry, node); + cxl_event_set_status(cxlds, log_type, true); + + /* Count went from 0 to 1 */ + return cxl_event_count(log) == 1; +} + +CXLRetCode cxl_event_get_records(CXLDeviceState *cxlds, CXLGetEventPayload *pl, + uint8_t log_type, int max_recs, + uint16_t *len) +{ + CXLEventLog *log; + CXLEvent *entry; + uint16_t nr; + + if (log_type >= CXL_EVENT_TYPE_MAX) { + return CXL_MBOX_INVALID_INPUT; + } + + log = &cxlds->event_logs[log_type]; + + QEMU_LOCK_GUARD(&log->lock); + + entry = cxl_event_get_head(log); + for (nr = 0; entry && nr < max_recs; nr++) { + memcpy(&pl->records[nr], &entry->data, CXL_EVENT_RECORD_SIZE); + entry = cxl_event_get_next(entry); + } + + if (!cxl_event_empty(log)) { + pl->flags |= CXL_GET_EVENT_FLAG_MORE_RECORDS; + } + + if (log->overflow_err_count) { + pl->flags |= CXL_GET_EVENT_FLAG_OVERFLOW; + pl->overflow_err_count = cpu_to_le16(log->overflow_err_count); + pl->first_overflow_timestamp = cpu_to_le64(log->first_overflow_timestamp); + pl->last_overflow_timestamp = cpu_to_le64(log->last_overflow_timestamp); + } + + pl->record_count = cpu_to_le16(nr); + *len = CXL_EVENT_PAYLOAD_HDR_SIZE + (CXL_EVENT_RECORD_SIZE * nr); + + return CXL_MBOX_SUCCESS; +} + +CXLRetCode cxl_event_clear_records(CXLDeviceState *cxlds, CXLClearEventPayload *pl) +{ + CXLEventLog *log; + uint8_t log_type; + CXLEvent *entry; + int nr; + + log_type = pl->event_log; + + if (log_type >= CXL_EVENT_TYPE_MAX) { + return CXL_MBOX_INVALID_INPUT; + } + + log = &cxlds->event_logs[log_type]; + + QEMU_LOCK_GUARD(&log->lock); + /* + * Must itterate the queue twice. + * "The device shall verify the event record handles specified in the input + * payload are in temporal order. If the device detects an older event + * record that will not be cleared when Clear Event Records is executed, + * the device shall return the Invalid Handle return code and shall not + * clear any of the specified event records." + * -- CXL 3.0 8.2.9.2.3 + */ + entry = cxl_event_get_head(log); + for (nr = 0; entry && nr < pl->nr_recs; nr++) { + uint16_t handle = pl->handle[nr]; + + /* NOTE: Both handles are little endian. */ + if (handle == 0 || entry->data.hdr.handle != handle) { + return CXL_MBOX_INVALID_INPUT; + } + entry = cxl_event_get_next(entry); + } + + entry = cxl_event_get_head(log); + for (nr = 0; entry && nr < pl->nr_recs; nr++) { + cxl_event_delete_head(cxlds, log_type, log); + entry = cxl_event_get_head(log); + } + + return CXL_MBOX_SUCCESS; +} diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c index d7e114aaae..3f46538048 100644 --- a/hw/cxl/cxl-mailbox-utils.c +++ b/hw/cxl/cxl-mailbox-utils.c @@ -9,6 +9,7 @@ #include "qemu/osdep.h" #include "hw/cxl/cxl.h" +#include "hw/cxl/cxl_events.h" #include "hw/pci/pci.h" #include "qemu/cutils.h" #include "qemu/log.h" @@ -95,11 +96,46 @@ struct cxl_cmd { return CXL_MBOX_SUCCESS; \ } -DEFINE_MAILBOX_HANDLER_ZEROED(events_get_records, 0x20); -DEFINE_MAILBOX_HANDLER_NOP(events_clear_records); DEFINE_MAILBOX_HANDLER_ZEROED(events_get_interrupt_policy, 4); DEFINE_MAILBOX_HANDLER_NOP(events_set_interrupt_policy); +static CXLRetCode cmd_events_get_records(struct cxl_cmd *cmd, + CXLDeviceState *cxlds, + uint16_t *len) +{ + CXLGetEventPayload *pl; + uint8_t log_type; + int max_recs; + + if (cmd->in < sizeof(log_type)) { + return CXL_MBOX_INVALID_INPUT; + } + + log_type = *((uint8_t *)cmd->payload); + + pl = (CXLGetEventPayload *)cmd->payload; + memset(pl, 0, sizeof(*pl)); + + max_recs = (cxlds->payload_size - CXL_EVENT_PAYLOAD_HDR_SIZE) / + CXL_EVENT_RECORD_SIZE; + if (max_recs > 0xFFFF) { + max_recs = 0xFFFF; + } + + return cxl_event_get_records(cxlds, pl, log_type, max_recs, len); +} + +static CXLRetCode cmd_events_clear_records(struct cxl_cmd *cmd, + CXLDeviceState *cxlds, + uint16_t *len) +{ + CXLClearEventPayload *pl; + + pl = (CXLClearEventPayload *)cmd->payload; + *len = 0; + return cxl_event_clear_records(cxlds, pl); +} + /* 8.2.9.2.1 */ static CXLRetCode cmd_firmware_update_get_info(struct cxl_cmd *cmd, CXLDeviceState *cxl_dstate, diff --git a/hw/cxl/meson.build b/hw/cxl/meson.build index 1f9aa2ea1f..e261ff3881 100644 --- a/hw/cxl/meson.build +++ b/hw/cxl/meson.build @@ -5,6 +5,7 @@ system_ss.add(when: 'CONFIG_CXL', 'cxl-mailbox-utils.c', 'cxl-host.c', 'cxl-cdat.c', + 'cxl-events.c', ), if_false: files( 'cxl-host-stubs.c', diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index d751803188..ec5a384885 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -724,6 +724,7 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp) goto err_release_cdat; } + cxl_event_init(&ct3d->cxl_dstate); return; err_release_cdat: diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h index 9f8ee85f8a..d3aec1bc0e 100644 --- a/include/hw/cxl/cxl_device.h +++ b/include/hw/cxl/cxl_device.h @@ -111,6 +111,20 @@ typedef enum { CXL_MBOX_MAX = 0x17 } CXLRetCode; +typedef struct CXLEvent { + CXLEventRecordRaw data; + QSIMPLEQ_ENTRY(CXLEvent) node; +} CXLEvent; + +typedef struct CXLEventLog { + uint16_t next_handle; + uint16_t overflow_err_count; + uint64_t first_overflow_timestamp; + uint64_t last_overflow_timestamp; + QemuMutex lock; + QSIMPLEQ_HEAD(, CXLEvent) events; +} CXLEventLog; + typedef struct cxl_device_state { MemoryRegion device_registers; @@ -161,6 +175,8 @@ typedef struct cxl_device_state { uint64_t mem_size; uint64_t pmem_size; uint64_t vmem_size; + + CXLEventLog event_logs[CXL_EVENT_TYPE_MAX]; } CXLDeviceState; /* Initialize the register block for a device */ @@ -353,6 +369,15 @@ MemTxResult cxl_type3_write(PCIDevice *d, hwaddr host_addr, uint64_t data, uint64_t cxl_device_get_timestamp(CXLDeviceState *cxlds); +void cxl_event_init(CXLDeviceState *cxlds); +bool cxl_event_insert(CXLDeviceState *cxlds, CXLEventLogType log_type, + CXLEventRecordRaw *event); +CXLRetCode cxl_event_get_records(CXLDeviceState *cxlds, CXLGetEventPayload *pl, + uint8_t log_type, int max_recs, + uint16_t *len); +CXLRetCode cxl_event_clear_records(CXLDeviceState *cxlds, + CXLClearEventPayload *pl); + void cxl_set_poison_list_overflowed(CXLType3Dev *ct3d); #endif diff --git a/include/hw/cxl/cxl_events.h b/include/hw/cxl/cxl_events.h index aeb3b0590e..d4aaa894f1 100644 --- a/include/hw/cxl/cxl_events.h +++ b/include/hw/cxl/cxl_events.h @@ -10,6 +10,8 @@ #ifndef CXL_EVENTS_H #define CXL_EVENTS_H +#include "qemu/uuid.h" + /* * CXL rev 3.0 section 8.2.9.2.2; Table 8-49 * @@ -25,4 +27,57 @@ typedef enum CXLEventLogType { CXL_EVENT_TYPE_MAX } CXLEventLogType; +/* + * Common Event Record Format + * CXL rev 3.0 section 8.2.9.2.1; Table 8-42 + */ +#define CXL_EVENT_REC_HDR_RES_LEN 0xf +typedef struct CXLEventRecordHdr { + QemuUUID id; + uint8_t length; + uint8_t flags[3]; + uint16_t handle; + uint16_t related_handle; + uint64_t timestamp; + uint8_t maint_op_class; + uint8_t reserved[CXL_EVENT_REC_HDR_RES_LEN]; +} QEMU_PACKED CXLEventRecordHdr; + +#define CXL_EVENT_RECORD_DATA_LENGTH 0x50 +typedef struct CXLEventRecordRaw { + CXLEventRecordHdr hdr; + uint8_t data[CXL_EVENT_RECORD_DATA_LENGTH]; +} QEMU_PACKED CXLEventRecordRaw; +#define CXL_EVENT_RECORD_SIZE (sizeof(CXLEventRecordRaw)) + +/* + * Get Event Records output payload + * CXL rev 3.0 section 8.2.9.2.2; Table 8-50 + */ +#define CXL_GET_EVENT_FLAG_OVERFLOW BIT(0) +#define CXL_GET_EVENT_FLAG_MORE_RECORDS BIT(1) +typedef struct CXLGetEventPayload { + uint8_t flags; + uint8_t reserved1; + uint16_t overflow_err_count; + uint64_t first_overflow_timestamp; + uint64_t last_overflow_timestamp; + uint16_t record_count; + uint8_t reserved2[0xa]; + CXLEventRecordRaw records[]; +} QEMU_PACKED CXLGetEventPayload; +#define CXL_EVENT_PAYLOAD_HDR_SIZE (sizeof(CXLGetEventPayload)) + +/* + * Clear Event Records input payload + * CXL rev 3.0 section 8.2.9.2.3; Table 8-51 + */ +typedef struct CXLClearEventPayload { + uint8_t event_log; /* CXLEventLogType */ + uint8_t clear_flags; + uint8_t nr_recs; + uint8_t reserved[3]; + uint16_t handle[]; +} CXLClearEventPayload; + #endif /* CXL_EVENTS_H */ -- cgit 1.4.1 From 6676bb973ba53d60886b06213ec98fbd736d66a1 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Tue, 30 May 2023 14:36:00 +0100 Subject: hw/cxl/events: Add event interrupt support Replace the stubbed out CXL Get/Set Event interrupt policy mailbox commands. Enable those commands to control interrupts for each of the event log types. Skip the standard input mailbox length on the Set command due to DCD being optional. Perform the checks separately. Signed-off-by: Ira Weiny Reviewed-by: Fan Ni Reviewed-by: Davidlohr Bueso Signed-off-by: Jonathan Cameron Message-Id: <20230530133603.16934-5-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/cxl/cxl-events.c | 33 +++++++++++++- hw/cxl/cxl-mailbox-utils.c | 106 +++++++++++++++++++++++++++++++++++--------- hw/mem/cxl_type3.c | 4 +- include/hw/cxl/cxl_device.h | 6 ++- include/hw/cxl/cxl_events.h | 23 ++++++++++ 5 files changed, 147 insertions(+), 25 deletions(-) (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/cxl/cxl-events.c b/hw/cxl/cxl-events.c index 5da1b76b97..d161d57456 100644 --- a/hw/cxl/cxl-events.c +++ b/hw/cxl/cxl-events.c @@ -13,6 +13,8 @@ #include "qemu/bswap.h" #include "qemu/typedefs.h" #include "qemu/error-report.h" +#include "hw/pci/msi.h" +#include "hw/pci/msix.h" #include "hw/cxl/cxl.h" #include "hw/cxl/cxl_events.h" @@ -26,7 +28,7 @@ static void reset_overflow(CXLEventLog *log) log->last_overflow_timestamp = 0; } -void cxl_event_init(CXLDeviceState *cxlds) +void cxl_event_init(CXLDeviceState *cxlds, int start_msg_num) { CXLEventLog *log; int i; @@ -37,9 +39,16 @@ void cxl_event_init(CXLDeviceState *cxlds) log->overflow_err_count = 0; log->first_overflow_timestamp = 0; log->last_overflow_timestamp = 0; + log->irq_enabled = false; + log->irq_vec = start_msg_num++; qemu_mutex_init(&log->lock); QSIMPLEQ_INIT(&log->events); } + + /* Override -- Dynamic Capacity uses the same vector as info */ + cxlds->event_logs[CXL_EVENT_TYPE_DYNAMIC_CAP].irq_vec = + cxlds->event_logs[CXL_EVENT_TYPE_INFO].irq_vec; + } static CXLEvent *cxl_event_get_head(CXLEventLog *log) @@ -215,3 +224,25 @@ CXLRetCode cxl_event_clear_records(CXLDeviceState *cxlds, CXLClearEventPayload * return CXL_MBOX_SUCCESS; } + +void cxl_event_irq_assert(CXLType3Dev *ct3d) +{ + CXLDeviceState *cxlds = &ct3d->cxl_dstate; + PCIDevice *pdev = &ct3d->parent_obj; + int i; + + for (i = 0; i < CXL_EVENT_TYPE_MAX; i++) { + CXLEventLog *log = &cxlds->event_logs[i]; + + if (!log->irq_enabled || cxl_event_empty(log)) { + continue; + } + + /* Notifies interrupt, legacy IRQ is not supported */ + if (msix_enabled(pdev)) { + msix_notify(pdev, log->irq_vec); + } else if (msi_enabled(pdev)) { + msi_notify(pdev, log->irq_vec); + } + } +} diff --git a/hw/cxl/cxl-mailbox-utils.c b/hw/cxl/cxl-mailbox-utils.c index 3f46538048..02f9b5a870 100644 --- a/hw/cxl/cxl-mailbox-utils.c +++ b/hw/cxl/cxl-mailbox-utils.c @@ -80,25 +80,6 @@ struct cxl_cmd { uint8_t *payload; }; -#define DEFINE_MAILBOX_HANDLER_ZEROED(name, size) \ - uint16_t __zero##name = size; \ - static CXLRetCode cmd_##name(struct cxl_cmd *cmd, \ - CXLDeviceState *cxl_dstate, uint16_t *len) \ - { \ - *len = __zero##name; \ - memset(cmd->payload, 0, *len); \ - return CXL_MBOX_SUCCESS; \ - } -#define DEFINE_MAILBOX_HANDLER_NOP(name) \ - static CXLRetCode cmd_##name(struct cxl_cmd *cmd, \ - CXLDeviceState *cxl_dstate, uint16_t *len) \ - { \ - return CXL_MBOX_SUCCESS; \ - } - -DEFINE_MAILBOX_HANDLER_ZEROED(events_get_interrupt_policy, 4); -DEFINE_MAILBOX_HANDLER_NOP(events_set_interrupt_policy); - static CXLRetCode cmd_events_get_records(struct cxl_cmd *cmd, CXLDeviceState *cxlds, uint16_t *len) @@ -136,6 +117,88 @@ static CXLRetCode cmd_events_clear_records(struct cxl_cmd *cmd, return cxl_event_clear_records(cxlds, pl); } +static CXLRetCode cmd_events_get_interrupt_policy(struct cxl_cmd *cmd, + CXLDeviceState *cxlds, + uint16_t *len) +{ + CXLEventInterruptPolicy *policy; + CXLEventLog *log; + + policy = (CXLEventInterruptPolicy *)cmd->payload; + memset(policy, 0, sizeof(*policy)); + + log = &cxlds->event_logs[CXL_EVENT_TYPE_INFO]; + if (log->irq_enabled) { + policy->info_settings = CXL_EVENT_INT_SETTING(log->irq_vec); + } + + log = &cxlds->event_logs[CXL_EVENT_TYPE_WARN]; + if (log->irq_enabled) { + policy->warn_settings = CXL_EVENT_INT_SETTING(log->irq_vec); + } + + log = &cxlds->event_logs[CXL_EVENT_TYPE_FAIL]; + if (log->irq_enabled) { + policy->failure_settings = CXL_EVENT_INT_SETTING(log->irq_vec); + } + + log = &cxlds->event_logs[CXL_EVENT_TYPE_FATAL]; + if (log->irq_enabled) { + policy->fatal_settings = CXL_EVENT_INT_SETTING(log->irq_vec); + } + + log = &cxlds->event_logs[CXL_EVENT_TYPE_DYNAMIC_CAP]; + if (log->irq_enabled) { + /* Dynamic Capacity borrows the same vector as info */ + policy->dyn_cap_settings = CXL_INT_MSI_MSIX; + } + + *len = sizeof(*policy); + return CXL_MBOX_SUCCESS; +} + +static CXLRetCode cmd_events_set_interrupt_policy(struct cxl_cmd *cmd, + CXLDeviceState *cxlds, + uint16_t *len) +{ + CXLEventInterruptPolicy *policy; + CXLEventLog *log; + + if (*len < CXL_EVENT_INT_SETTING_MIN_LEN) { + return CXL_MBOX_INVALID_PAYLOAD_LENGTH; + } + + policy = (CXLEventInterruptPolicy *)cmd->payload; + + log = &cxlds->event_logs[CXL_EVENT_TYPE_INFO]; + log->irq_enabled = (policy->info_settings & CXL_EVENT_INT_MODE_MASK) == + CXL_INT_MSI_MSIX; + + log = &cxlds->event_logs[CXL_EVENT_TYPE_WARN]; + log->irq_enabled = (policy->warn_settings & CXL_EVENT_INT_MODE_MASK) == + CXL_INT_MSI_MSIX; + + log = &cxlds->event_logs[CXL_EVENT_TYPE_FAIL]; + log->irq_enabled = (policy->failure_settings & CXL_EVENT_INT_MODE_MASK) == + CXL_INT_MSI_MSIX; + + log = &cxlds->event_logs[CXL_EVENT_TYPE_FATAL]; + log->irq_enabled = (policy->fatal_settings & CXL_EVENT_INT_MODE_MASK) == + CXL_INT_MSI_MSIX; + + /* DCD is optional */ + if (*len < sizeof(*policy)) { + return CXL_MBOX_SUCCESS; + } + + log = &cxlds->event_logs[CXL_EVENT_TYPE_DYNAMIC_CAP]; + log->irq_enabled = (policy->dyn_cap_settings & CXL_EVENT_INT_MODE_MASK) == + CXL_INT_MSI_MSIX; + + *len = sizeof(*policy); + return CXL_MBOX_SUCCESS; +} + /* 8.2.9.2.1 */ static CXLRetCode cmd_firmware_update_get_info(struct cxl_cmd *cmd, CXLDeviceState *cxl_dstate, @@ -611,9 +674,10 @@ static struct cxl_cmd cxl_cmd_set[256][256] = { [EVENTS][CLEAR_RECORDS] = { "EVENTS_CLEAR_RECORDS", cmd_events_clear_records, ~0, IMMEDIATE_LOG_CHANGE }, [EVENTS][GET_INTERRUPT_POLICY] = { "EVENTS_GET_INTERRUPT_POLICY", - cmd_events_get_interrupt_policy, 0, 0 }, + cmd_events_get_interrupt_policy, 0, 0 }, [EVENTS][SET_INTERRUPT_POLICY] = { "EVENTS_SET_INTERRUPT_POLICY", - cmd_events_set_interrupt_policy, 4, IMMEDIATE_CONFIG_CHANGE }, + cmd_events_set_interrupt_policy, + ~0, IMMEDIATE_CONFIG_CHANGE }, [FIRMWARE_UPDATE][GET_INFO] = { "FIRMWARE_UPDATE_GET_INFO", cmd_firmware_update_get_info, 0, 0 }, [TIMESTAMP][GET] = { "TIMESTAMP_GET", cmd_timestamp_get, 0, 0 }, diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index ec5a384885..c9e347f42b 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -659,7 +659,7 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp) ComponentRegisters *regs = &cxl_cstate->crb; MemoryRegion *mr = ®s->component_registers; uint8_t *pci_conf = pci_dev->config; - unsigned short msix_num = 1; + unsigned short msix_num = 6; int i, rc; QTAILQ_INIT(&ct3d->error_list); @@ -723,8 +723,8 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp) if (rc) { goto err_release_cdat; } + cxl_event_init(&ct3d->cxl_dstate, 2); - cxl_event_init(&ct3d->cxl_dstate); return; err_release_cdat: diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h index d3aec1bc0e..1978730fba 100644 --- a/include/hw/cxl/cxl_device.h +++ b/include/hw/cxl/cxl_device.h @@ -121,6 +121,8 @@ typedef struct CXLEventLog { uint16_t overflow_err_count; uint64_t first_overflow_timestamp; uint64_t last_overflow_timestamp; + bool irq_enabled; + int irq_vec; QemuMutex lock; QSIMPLEQ_HEAD(, CXLEvent) events; } CXLEventLog; @@ -369,7 +371,7 @@ MemTxResult cxl_type3_write(PCIDevice *d, hwaddr host_addr, uint64_t data, uint64_t cxl_device_get_timestamp(CXLDeviceState *cxlds); -void cxl_event_init(CXLDeviceState *cxlds); +void cxl_event_init(CXLDeviceState *cxlds, int start_msg_num); bool cxl_event_insert(CXLDeviceState *cxlds, CXLEventLogType log_type, CXLEventRecordRaw *event); CXLRetCode cxl_event_get_records(CXLDeviceState *cxlds, CXLGetEventPayload *pl, @@ -378,6 +380,8 @@ CXLRetCode cxl_event_get_records(CXLDeviceState *cxlds, CXLGetEventPayload *pl, CXLRetCode cxl_event_clear_records(CXLDeviceState *cxlds, CXLClearEventPayload *pl); +void cxl_event_irq_assert(CXLType3Dev *ct3d); + void cxl_set_poison_list_overflowed(CXLType3Dev *ct3d); #endif diff --git a/include/hw/cxl/cxl_events.h b/include/hw/cxl/cxl_events.h index d4aaa894f1..4bf8b7aa08 100644 --- a/include/hw/cxl/cxl_events.h +++ b/include/hw/cxl/cxl_events.h @@ -80,4 +80,27 @@ typedef struct CXLClearEventPayload { uint16_t handle[]; } CXLClearEventPayload; +/** + * Event Interrupt Policy + * + * CXL rev 3.0 section 8.2.9.2.4; Table 8-52 + */ +typedef enum CXLEventIntMode { + CXL_INT_NONE = 0x00, + CXL_INT_MSI_MSIX = 0x01, + CXL_INT_FW = 0x02, + CXL_INT_RES = 0x03, +} CXLEventIntMode; +#define CXL_EVENT_INT_MODE_MASK 0x3 +#define CXL_EVENT_INT_SETTING(vector) ((((uint8_t)vector & 0xf) << 4) | CXL_INT_MSI_MSIX) +typedef struct CXLEventInterruptPolicy { + uint8_t info_settings; + uint8_t warn_settings; + uint8_t failure_settings; + uint8_t fatal_settings; + uint8_t dyn_cap_settings; +} QEMU_PACKED CXLEventInterruptPolicy; +/* DCD is optional but other fields are not */ +#define CXL_EVENT_INT_SETTING_MIN_LEN 4 + #endif /* CXL_EVENTS_H */ -- cgit 1.4.1 From ea9b6d647f2f4708708d19ba1cb17d332d3eff06 Mon Sep 17 00:00:00 2001 From: Ira Weiny Date: Tue, 30 May 2023 14:36:01 +0100 Subject: hw/cxl/events: Add injection of General Media Events To facilitate testing provide a QMP command to inject a general media event. The event can be added to the log specified. Signed-off-by: Ira Weiny Reviewed-by: Fan Ni Acked-by: Markus Armbruster Signed-off-by: Jonathan Cameron Message-Id: <20230530133603.16934-6-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/mem/cxl_type3.c | 111 ++++++++++++++++++++++++++++++++++++++++++++ hw/mem/cxl_type3_stubs.c | 10 ++++ include/hw/cxl/cxl_events.h | 20 ++++++++ qapi/cxl.json | 74 +++++++++++++++++++++++++++++ 4 files changed, 215 insertions(+) (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index c9e347f42b..b1618779d2 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -1181,6 +1181,117 @@ void qmp_cxl_inject_correctable_error(const char *path, CxlCorErrorType type, pcie_aer_inject_error(PCI_DEVICE(obj), &err); } +static void cxl_assign_event_header(CXLEventRecordHdr *hdr, + const QemuUUID *uuid, uint32_t flags, + uint8_t length, uint64_t timestamp) +{ + st24_le_p(&hdr->flags, flags); + hdr->length = length; + memcpy(&hdr->id, uuid, sizeof(hdr->id)); + stq_le_p(&hdr->timestamp, timestamp); +} + +static const QemuUUID gen_media_uuid = { + .data = UUID(0xfbcd0a77, 0xc260, 0x417f, + 0x85, 0xa9, 0x08, 0x8b, 0x16, 0x21, 0xeb, 0xa6), +}; + +#define CXL_GMER_VALID_CHANNEL BIT(0) +#define CXL_GMER_VALID_RANK BIT(1) +#define CXL_GMER_VALID_DEVICE BIT(2) +#define CXL_GMER_VALID_COMPONENT BIT(3) + +static int ct3d_qmp_cxl_event_log_enc(CxlEventLog log) +{ + switch (log) { + case CXL_EVENT_LOG_INFORMATIONAL: + return CXL_EVENT_TYPE_INFO; + case CXL_EVENT_LOG_WARNING: + return CXL_EVENT_TYPE_WARN; + case CXL_EVENT_LOG_FAILURE: + return CXL_EVENT_TYPE_FAIL; + case CXL_EVENT_LOG_FATAL: + return CXL_EVENT_TYPE_FATAL; +/* DCD not yet supported */ + default: + return -EINVAL; + } +} +/* Component ID is device specific. Define this as a string. */ +void qmp_cxl_inject_general_media_event(const char *path, CxlEventLog log, + uint8_t flags, uint64_t dpa, + uint8_t descriptor, uint8_t type, + uint8_t transaction_type, + bool has_channel, uint8_t channel, + bool has_rank, uint8_t rank, + bool has_device, uint32_t device, + const char *component_id, + Error **errp) +{ + Object *obj = object_resolve_path(path, NULL); + CXLEventGenMedia gem; + CXLEventRecordHdr *hdr = &gem.hdr; + CXLDeviceState *cxlds; + CXLType3Dev *ct3d; + uint16_t valid_flags = 0; + uint8_t enc_log; + int rc; + + if (!obj) { + error_setg(errp, "Unable to resolve path"); + return; + } + if (!object_dynamic_cast(obj, TYPE_CXL_TYPE3)) { + error_setg(errp, "Path does not point to a CXL type 3 device"); + return; + } + ct3d = CXL_TYPE3(obj); + cxlds = &ct3d->cxl_dstate; + + rc = ct3d_qmp_cxl_event_log_enc(log); + if (rc < 0) { + error_setg(errp, "Unhandled error log type"); + return; + } + enc_log = rc; + + memset(&gem, 0, sizeof(gem)); + cxl_assign_event_header(hdr, &gen_media_uuid, flags, sizeof(gem), + cxl_device_get_timestamp(&ct3d->cxl_dstate)); + + stq_le_p(&gem.phys_addr, dpa); + gem.descriptor = descriptor; + gem.type = type; + gem.transaction_type = transaction_type; + + if (has_channel) { + gem.channel = channel; + valid_flags |= CXL_GMER_VALID_CHANNEL; + } + + if (has_rank) { + gem.rank = rank; + valid_flags |= CXL_GMER_VALID_RANK; + } + + if (has_device) { + st24_le_p(gem.device, device); + valid_flags |= CXL_GMER_VALID_DEVICE; + } + + if (component_id) { + strncpy((char *)gem.component_id, component_id, + sizeof(gem.component_id) - 1); + valid_flags |= CXL_GMER_VALID_COMPONENT; + } + + stw_le_p(&gem.validity_flags, valid_flags); + + if (cxl_event_insert(cxlds, enc_log, (CXLEventRecordRaw *)&gem)) { + cxl_event_irq_assert(ct3d); + } +} + static void ct3_class_init(ObjectClass *oc, void *data) { DeviceClass *dc = DEVICE_CLASS(oc); diff --git a/hw/mem/cxl_type3_stubs.c b/hw/mem/cxl_type3_stubs.c index fd1166a610..4dfbdf9268 100644 --- a/hw/mem/cxl_type3_stubs.c +++ b/hw/mem/cxl_type3_stubs.c @@ -3,6 +3,16 @@ #include "qapi/error.h" #include "qapi/qapi-commands-cxl.h" +void qmp_cxl_inject_general_media_event(const char *path, CxlEventLog log, + uint8_t flags, uint64_t dpa, + uint8_t descriptor, uint8_t type, + uint8_t transaction_type, + bool has_channel, uint8_t channel, + bool has_rank, uint8_t rank, + bool has_device, uint32_t device, + const char *component_id, + Error **errp) {} + void qmp_cxl_inject_poison(const char *path, uint64_t start, uint64_t length, Error **errp) { diff --git a/include/hw/cxl/cxl_events.h b/include/hw/cxl/cxl_events.h index 4bf8b7aa08..b189193f4c 100644 --- a/include/hw/cxl/cxl_events.h +++ b/include/hw/cxl/cxl_events.h @@ -103,4 +103,24 @@ typedef struct CXLEventInterruptPolicy { /* DCD is optional but other fields are not */ #define CXL_EVENT_INT_SETTING_MIN_LEN 4 +/* + * General Media Event Record + * CXL rev 3.0 Section 8.2.9.2.1.1; Table 8-43 + */ +#define CXL_EVENT_GEN_MED_COMP_ID_SIZE 0x10 +#define CXL_EVENT_GEN_MED_RES_SIZE 0x2e +typedef struct CXLEventGenMedia { + CXLEventRecordHdr hdr; + uint64_t phys_addr; + uint8_t descriptor; + uint8_t type; + uint8_t transaction_type; + uint16_t validity_flags; + uint8_t channel; + uint8_t rank; + uint8_t device[3]; + uint8_t component_id[CXL_EVENT_GEN_MED_COMP_ID_SIZE]; + uint8_t reserved[CXL_EVENT_GEN_MED_RES_SIZE]; +} QEMU_PACKED CXLEventGenMedia; + #endif /* CXL_EVENTS_H */ diff --git a/qapi/cxl.json b/qapi/cxl.json index ed1c7eea3a..d509430844 100644 --- a/qapi/cxl.json +++ b/qapi/cxl.json @@ -5,6 +5,80 @@ # = CXL devices ## +## +# @CxlEventLog: +# +# CXL has a number of separate event logs for different types of +# events. Each such event log is handled and signaled independently. +# +# @informational: Information Event Log +# +# @warning: Warning Event Log +# +# @failure: Failure Event Log +# +# @fatal: Fatal Event Log +# +# Since: 8.1 +## +{ 'enum': 'CxlEventLog', + 'data': ['informational', + 'warning', + 'failure', + 'fatal'] + } + +## +# @cxl-inject-general-media-event: +# +# Inject an event record for a General Media Event (CXL r3.0 +# 8.2.9.2.1.1). This event type is reported via one of the event logs +# specified via the log parameter. +# +# @path: CXL type 3 device canonical QOM path +# +# @log: event log to add the event to +# +# @flags: Event Record Flags. See CXL r3.0 Table 8-42 Common Event +# Record Format, Event Record Flags for subfield definitions. +# +# @dpa: Device Physical Address (relative to @path device). Note +# lower bits include some flags. See CXL r3.0 Table 8-43 General +# Media Event Record, Physical Address. +# +# @descriptor: Memory Event Descriptor with additional memory event +# information. See CXL r3.0 Table 8-43 General Media Event +# Record, Memory Event Descriptor for bit definitions. +# +# @type: Type of memory event that occurred. See CXL r3.0 Table 8-43 +# General Media Event Record, Memory Event Type for possible +# values. +# +# @transaction-type: Type of first transaction that caused the event +# to occur. See CXL r3.0 Table 8-43 General Media Event Record, +# Transaction Type for possible values. +# +# @channel: The channel of the memory event location. A channel is an +# interface that can be independently accessed for a transaction. +# +# @rank: The rank of the memory event location. A rank is a set of +# memory devices on a channel that together execute a transaction. +# +# @device: Bitmask that represents all devices in the rank associated +# with the memory event location. +# +# @component-id: Device specific component identifier for the event. +# May describe a field replaceable sub-component of the device. +# +# Since: 8.1 +## +{ 'command': 'cxl-inject-general-media-event', + 'data': { 'path': 'str', 'log': 'CxlEventLog', 'flags': 'uint8', + 'dpa': 'uint64', 'descriptor': 'uint8', + 'type': 'uint8', 'transaction-type': 'uint8', + '*channel': 'uint8', '*rank': 'uint8', + '*device': 'uint32', '*component-id': 'str' } } + ## # @cxl-inject-poison: # -- cgit 1.4.1 From b90a324eda7113b62b558aad43e2166eb52567d2 Mon Sep 17 00:00:00 2001 From: Jonathan Cameron Date: Tue, 30 May 2023 14:36:02 +0100 Subject: hw/cxl/events: Add injection of DRAM events Defined in CXL r3.0 8.2.9.2.1.2 DRAM Event Record, this event provides information related to DRAM devices. Example injection command in QMP: { "execute": "cxl-inject-dram-event", "arguments": { "path": "/machine/peripheral/cxl-mem0", "log": "informational", "flags": 1, "dpa": 1000, "descriptor": 3, "type": 3, "transaction-type": 192, "channel": 3, "rank": 17, "nibble-mask": 37421234, "bank-group": 7, "bank": 11, "row": 2, "column": 77, "correction-mask": [33, 44, 55,66] }} Acked-by: Markus Armbruster Reviewed-by: Fan Ni Reviewed-by: Ira Weiny Signed-off-by: Jonathan Cameron Message-Id: <20230530133603.16934-7-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/mem/cxl_type3.c | 116 ++++++++++++++++++++++++++++++++++++++++++++ hw/mem/cxl_type3_stubs.c | 13 +++++ include/hw/cxl/cxl_events.h | 23 +++++++++ qapi/cxl.json | 61 +++++++++++++++++++++++ 4 files changed, 213 insertions(+) (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index b1618779d2..3c07b1b7a3 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -1196,6 +1196,11 @@ static const QemuUUID gen_media_uuid = { 0x85, 0xa9, 0x08, 0x8b, 0x16, 0x21, 0xeb, 0xa6), }; +static const QemuUUID dram_uuid = { + .data = UUID(0x601dcbb3, 0x9c06, 0x4eab, 0xb8, 0xaf, + 0x4e, 0x9b, 0xfb, 0x5c, 0x96, 0x24), +}; + #define CXL_GMER_VALID_CHANNEL BIT(0) #define CXL_GMER_VALID_RANK BIT(1) #define CXL_GMER_VALID_DEVICE BIT(2) @@ -1292,6 +1297,117 @@ void qmp_cxl_inject_general_media_event(const char *path, CxlEventLog log, } } +#define CXL_DRAM_VALID_CHANNEL BIT(0) +#define CXL_DRAM_VALID_RANK BIT(1) +#define CXL_DRAM_VALID_NIBBLE_MASK BIT(2) +#define CXL_DRAM_VALID_BANK_GROUP BIT(3) +#define CXL_DRAM_VALID_BANK BIT(4) +#define CXL_DRAM_VALID_ROW BIT(5) +#define CXL_DRAM_VALID_COLUMN BIT(6) +#define CXL_DRAM_VALID_CORRECTION_MASK BIT(7) + +void qmp_cxl_inject_dram_event(const char *path, CxlEventLog log, uint8_t flags, + uint64_t dpa, uint8_t descriptor, + uint8_t type, uint8_t transaction_type, + bool has_channel, uint8_t channel, + bool has_rank, uint8_t rank, + bool has_nibble_mask, uint32_t nibble_mask, + bool has_bank_group, uint8_t bank_group, + bool has_bank, uint8_t bank, + bool has_row, uint32_t row, + bool has_column, uint16_t column, + bool has_correction_mask, uint64List *correction_mask, + Error **errp) +{ + Object *obj = object_resolve_path(path, NULL); + CXLEventDram dram; + CXLEventRecordHdr *hdr = &dram.hdr; + CXLDeviceState *cxlds; + CXLType3Dev *ct3d; + uint16_t valid_flags = 0; + uint8_t enc_log; + int rc; + + if (!obj) { + error_setg(errp, "Unable to resolve path"); + return; + } + if (!object_dynamic_cast(obj, TYPE_CXL_TYPE3)) { + error_setg(errp, "Path does not point to a CXL type 3 device"); + return; + } + ct3d = CXL_TYPE3(obj); + cxlds = &ct3d->cxl_dstate; + + rc = ct3d_qmp_cxl_event_log_enc(log); + if (rc < 0) { + error_setg(errp, "Unhandled error log type"); + return; + } + enc_log = rc; + + memset(&dram, 0, sizeof(dram)); + cxl_assign_event_header(hdr, &dram_uuid, flags, sizeof(dram), + cxl_device_get_timestamp(&ct3d->cxl_dstate)); + stq_le_p(&dram.phys_addr, dpa); + dram.descriptor = descriptor; + dram.type = type; + dram.transaction_type = transaction_type; + + if (has_channel) { + dram.channel = channel; + valid_flags |= CXL_DRAM_VALID_CHANNEL; + } + + if (has_rank) { + dram.rank = rank; + valid_flags |= CXL_DRAM_VALID_RANK; + } + + if (has_nibble_mask) { + st24_le_p(dram.nibble_mask, nibble_mask); + valid_flags |= CXL_DRAM_VALID_NIBBLE_MASK; + } + + if (has_bank_group) { + dram.bank_group = bank_group; + valid_flags |= CXL_DRAM_VALID_BANK_GROUP; + } + + if (has_bank) { + dram.bank = bank; + valid_flags |= CXL_DRAM_VALID_BANK; + } + + if (has_row) { + st24_le_p(dram.row, row); + valid_flags |= CXL_DRAM_VALID_ROW; + } + + if (has_column) { + stw_le_p(&dram.column, column); + valid_flags |= CXL_DRAM_VALID_COLUMN; + } + + if (has_correction_mask) { + int count = 0; + while (correction_mask && count < 4) { + stq_le_p(&dram.correction_mask[count], + correction_mask->value); + count++; + correction_mask = correction_mask->next; + } + valid_flags |= CXL_DRAM_VALID_CORRECTION_MASK; + } + + stw_le_p(&dram.validity_flags, valid_flags); + + if (cxl_event_insert(cxlds, enc_log, (CXLEventRecordRaw *)&dram)) { + cxl_event_irq_assert(ct3d); + } + return; +} + static void ct3_class_init(ObjectClass *oc, void *data) { DeviceClass *dc = DEVICE_CLASS(oc); diff --git a/hw/mem/cxl_type3_stubs.c b/hw/mem/cxl_type3_stubs.c index 4dfbdf9268..e904c5d089 100644 --- a/hw/mem/cxl_type3_stubs.c +++ b/hw/mem/cxl_type3_stubs.c @@ -13,6 +13,19 @@ void qmp_cxl_inject_general_media_event(const char *path, CxlEventLog log, const char *component_id, Error **errp) {} +void qmp_cxl_inject_dram_event(const char *path, CxlEventLog log, uint8_t flags, + uint64_t dpa, uint8_t descriptor, + uint8_t type, uint8_t transaction_type, + bool has_channel, uint8_t channel, + bool has_rank, uint8_t rank, + bool has_nibble_mask, uint32_t nibble_mask, + bool has_bank_group, uint8_t bank_group, + bool has_bank, uint8_t bank, + bool has_row, uint32_t row, + bool has_column, uint16_t column, + bool has_correction_mask, uint64List *correction_mask, + Error **errp) {} + void qmp_cxl_inject_poison(const char *path, uint64_t start, uint64_t length, Error **errp) { diff --git a/include/hw/cxl/cxl_events.h b/include/hw/cxl/cxl_events.h index b189193f4c..a39e30d973 100644 --- a/include/hw/cxl/cxl_events.h +++ b/include/hw/cxl/cxl_events.h @@ -123,4 +123,27 @@ typedef struct CXLEventGenMedia { uint8_t reserved[CXL_EVENT_GEN_MED_RES_SIZE]; } QEMU_PACKED CXLEventGenMedia; +/* + * DRAM Event Record + * CXL Rev 3.0 Section 8.2.9.2.1.2: Table 8-44 + * All fields little endian. + */ +typedef struct CXLEventDram { + CXLEventRecordHdr hdr; + uint64_t phys_addr; + uint8_t descriptor; + uint8_t type; + uint8_t transaction_type; + uint16_t validity_flags; + uint8_t channel; + uint8_t rank; + uint8_t nibble_mask[3]; + uint8_t bank_group; + uint8_t bank; + uint8_t row[3]; + uint16_t column; + uint64_t correction_mask[4]; + uint8_t reserved[0x17]; +} QEMU_PACKED CXLEventDram; + #endif /* CXL_EVENTS_H */ diff --git a/qapi/cxl.json b/qapi/cxl.json index d509430844..2ad310387c 100644 --- a/qapi/cxl.json +++ b/qapi/cxl.json @@ -79,6 +79,67 @@ '*channel': 'uint8', '*rank': 'uint8', '*device': 'uint32', '*component-id': 'str' } } +## +# @cxl-inject-dram-event: +# +# Inject an event record for a DRAM Event (CXL r3.0 8.2.9.2.1.2). +# This event type is reported via one of the event logs specified via +# the log parameter. +# +# @path: CXL type 3 device canonical QOM path +# +# @log: Event log to add the event to +# +# @flags: Event Record Flags. See CXL r3.0 Table 8-42 Common Event +# Record Format, Event Record Flags for subfield definitions. +# +# @dpa: Device Physical Address (relative to @path device). Note +# lower bits include some flags. See CXL r3.0 Table 8-44 DRAM +# Event Record, Physical Address. +# +# @descriptor: Memory Event Descriptor with additional memory event +# information. See CXL r3.0 Table 8-44 DRAM Event Record, Memory +# Event Descriptor for bit definitions. +# +# @type: Type of memory event that occurred. See CXL r3.0 Table 8-44 +# DRAM Event Record, Memory Event Type for possible values. +# +# @transaction-type: Type of first transaction that caused the event +# to occur. See CXL r3.0 Table 8-44 DRAM Event Record, +# Transaction Type for possible values. +# +# @channel: The channel of the memory event location. A channel is an +# interface that can be independently accessed for a transaction. +# +# @rank: The rank of the memory event location. A rank is a set of +# memory devices on a channel that together execute a transaction. +# +# @nibble-mask: Identifies one or more nibbles that the error affects +# +# @bank-group: Bank group of the memory event location, incorporating +# a number of Banks. +# +# @bank: Bank of the memory event location. A single bank is accessed +# per read or write of the memory. +# +# @row: Row address within the DRAM. +# +# @column: Column address within the DRAM. +# +# @correction-mask: Bits within each nibble. Used in order of bits +# set in the nibble-mask. Up to 4 nibbles may be covered. +# +# Since: 8.1 +## +{ 'command': 'cxl-inject-dram-event', + 'data': { 'path': 'str', 'log': 'CxlEventLog', 'flags': 'uint8', + 'dpa': 'uint64', 'descriptor': 'uint8', + 'type': 'uint8', 'transaction-type': 'uint8', + '*channel': 'uint8', '*rank': 'uint8', '*nibble-mask': 'uint32', + '*bank-group': 'uint8', '*bank': 'uint8', '*row': 'uint32', + '*column': 'uint16', '*correction-mask': [ 'uint64' ] + }} + ## # @cxl-inject-poison: # -- cgit 1.4.1 From bafe03083255da3a053144b77a5fbc7dbf9494f3 Mon Sep 17 00:00:00 2001 From: Jonathan Cameron Date: Tue, 30 May 2023 14:36:03 +0100 Subject: hw/cxl/events: Add injection of Memory Module Events These events include a copy of the device health information at the time of the event. Actually using the emulated device health would require a lot of controls to manipulate that state. Given the aim of this injection code is to just test the flows when events occur, inject the contents of the device health state as well. Future work may add more sophisticate device health emulation including direct generation of these records when events occur (such as a temperature threshold being crossed). That does not reduce the usefulness of this more basic generation of the events. Acked-by: Markus Armbruster Reviewed-by: Fan Ni Reviewed-by: Ira Weiny Signed-off-by: Jonathan Cameron Message-Id: <20230530133603.16934-8-Jonathan.Cameron@huawei.com> Reviewed-by: Michael S. Tsirkin Signed-off-by: Michael S. Tsirkin --- hw/mem/cxl_type3.c | 62 +++++++++++++++++++++++++++++++++++++++++++++ hw/mem/cxl_type3_stubs.c | 12 +++++++++ include/hw/cxl/cxl_events.h | 19 ++++++++++++++ qapi/cxl.json | 53 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 146 insertions(+) (limited to 'hw/mem/cxl_type3.c') diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c index 3c07b1b7a3..4e314748d3 100644 --- a/hw/mem/cxl_type3.c +++ b/hw/mem/cxl_type3.c @@ -1201,6 +1201,11 @@ static const QemuUUID dram_uuid = { 0x4e, 0x9b, 0xfb, 0x5c, 0x96, 0x24), }; +static const QemuUUID memory_module_uuid = { + .data = UUID(0xfe927475, 0xdd59, 0x4339, 0xa5, 0x86, + 0x79, 0xba, 0xb1, 0x13, 0xb7, 0x74), +}; + #define CXL_GMER_VALID_CHANNEL BIT(0) #define CXL_GMER_VALID_RANK BIT(1) #define CXL_GMER_VALID_DEVICE BIT(2) @@ -1408,6 +1413,63 @@ void qmp_cxl_inject_dram_event(const char *path, CxlEventLog log, uint8_t flags, return; } +void qmp_cxl_inject_memory_module_event(const char *path, CxlEventLog log, + uint8_t flags, uint8_t type, + uint8_t health_status, + uint8_t media_status, + uint8_t additional_status, + uint8_t life_used, + int16_t temperature, + uint32_t dirty_shutdown_count, + uint32_t corrected_volatile_error_count, + uint32_t corrected_persistent_error_count, + Error **errp) +{ + Object *obj = object_resolve_path(path, NULL); + CXLEventMemoryModule module; + CXLEventRecordHdr *hdr = &module.hdr; + CXLDeviceState *cxlds; + CXLType3Dev *ct3d; + uint8_t enc_log; + int rc; + + if (!obj) { + error_setg(errp, "Unable to resolve path"); + return; + } + if (!object_dynamic_cast(obj, TYPE_CXL_TYPE3)) { + error_setg(errp, "Path does not point to a CXL type 3 device"); + return; + } + ct3d = CXL_TYPE3(obj); + cxlds = &ct3d->cxl_dstate; + + rc = ct3d_qmp_cxl_event_log_enc(log); + if (rc < 0) { + error_setg(errp, "Unhandled error log type"); + return; + } + enc_log = rc; + + memset(&module, 0, sizeof(module)); + cxl_assign_event_header(hdr, &memory_module_uuid, flags, sizeof(module), + cxl_device_get_timestamp(&ct3d->cxl_dstate)); + + module.type = type; + module.health_status = health_status; + module.media_status = media_status; + module.additional_status = additional_status; + module.life_used = life_used; + stw_le_p(&module.temperature, temperature); + stl_le_p(&module.dirty_shutdown_count, dirty_shutdown_count); + stl_le_p(&module.corrected_volatile_error_count, corrected_volatile_error_count); + stl_le_p(&module.corrected_persistent_error_count, corrected_persistent_error_count); + + if (cxl_event_insert(cxlds, enc_log, (CXLEventRecordRaw *)&module)) { + cxl_event_irq_assert(ct3d); + } +} + static void ct3_class_init(ObjectClass *oc, void *data) { DeviceClass *dc = DEVICE_CLASS(oc); diff --git a/hw/mem/cxl_type3_stubs.c b/hw/mem/cxl_type3_stubs.c index e904c5d089..f3e4a9fa72 100644 --- a/hw/mem/cxl_type3_stubs.c +++ b/hw/mem/cxl_type3_stubs.c @@ -26,6 +26,18 @@ void qmp_cxl_inject_dram_event(const char *path, CxlEventLog log, uint8_t flags, bool has_correction_mask, uint64List *correction_mask, Error **errp) {} +void qmp_cxl_inject_memory_module_event(const char *path, CxlEventLog log, + uint8_t flags, uint8_t type, + uint8_t health_status, + uint8_t media_status, + uint8_t additional_status, + uint8_t life_used, + int16_t temperature, + uint32_t dirty_shutdown_count, + uint32_t corrected_volatile_error_count, + uint32_t corrected_persistent_error_count, + Error **errp) {} + void qmp_cxl_inject_poison(const char *path, uint64_t start, uint64_t length, Error **errp) { diff --git a/include/hw/cxl/cxl_events.h b/include/hw/cxl/cxl_events.h index a39e30d973..089ba2091f 100644 --- a/include/hw/cxl/cxl_events.h +++ b/include/hw/cxl/cxl_events.h @@ -146,4 +146,23 @@ typedef struct CXLEventDram { uint8_t reserved[0x17]; } QEMU_PACKED CXLEventDram; +/* + * Memory Module Event Record + * CXL Rev 3.0 Section 8.2.9.2.1.3: Table 8-45 + * All fields little endian. + */ +typedef struct CXLEventMemoryModule { + CXLEventRecordHdr hdr; + uint8_t type; + uint8_t health_status; + uint8_t media_status; + uint8_t additional_status; + uint8_t life_used; + int16_t temperature; + uint32_t dirty_shutdown_count; + uint32_t corrected_volatile_error_count; + uint32_t corrected_persistent_error_count; + uint8_t reserved[0x3d]; +} QEMU_PACKED CXLEventMemoryModule; + #endif /* CXL_EVENTS_H */ diff --git a/qapi/cxl.json b/qapi/cxl.json index 2ad310387c..d5b5293eb5 100644 --- a/qapi/cxl.json +++ b/qapi/cxl.json @@ -140,6 +140,59 @@ '*column': 'uint16', '*correction-mask': [ 'uint64' ] }} +## +# @cxl-inject-memory-module-event: +# +# Inject an event record for a Memory Module Event (CXL r3.0 +# 8.2.9.2.1.3). This event includes a copy of the Device Health +# info at the time of the event. +# +# @path: CXL type 3 device canonical QOM path +# +# @log: Event Log to add the event to +# +# @flags: Event Record Flags. See CXL r3.0 Table 8-42 Common Event +# Record Format, Event Record Flags for subfield definitions. +# +# @type: Device Event Type. See CXL r3.0 Table 8-45 Memory Module +# Event Record for bit definitions for bit definiions. +# +# @health-status: Overall health summary bitmap. See CXL r3.0 Table +# 8-100 Get Health Info Output Payload, Health Status for bit +# definitions. +# +# @media-status: Overall media health summary. See CXL r3.0 Table +# 8-100 Get Health Info Output Payload, Media Status for bit +# definitions. +# +# @additional-status: See CXL r3.0 Table 8-100 Get Health Info Output +# Payload, Additional Status for subfield definitions. +# +# @life-used: Percentage (0-100) of factory expected life span. +# +# @temperature: Device temperature in degrees Celsius. +# +# @dirty-shutdown-count: Number of times the device has been unable +# to determine whether data loss may have occurred. +# +# @corrected-volatile-error-count: Total number of correctable errors +# in volatile memory. +# +# @corrected-persistent-error-count: Total number of correctable +# errors in persistent memory +# +# Since: 8.1 +## +{ 'command': 'cxl-inject-memory-module-event', + 'data': { 'path': 'str', 'log': 'CxlEventLog', 'flags' : 'uint8', + 'type': 'uint8', 'health-status': 'uint8', + 'media-status': 'uint8', 'additional-status': 'uint8', + 'life-used': 'uint8', 'temperature' : 'int16', + 'dirty-shutdown-count': 'uint32', + 'corrected-volatile-error-count': 'uint32', + 'corrected-persistent-error-count': 'uint32' + }} + ## # @cxl-inject-poison: # -- cgit 1.4.1