| Message ID | 20220906220552.1243998-2-titusr@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Initial PECI bus support | expand |
On Tue, Sep 06, 2022 at 10:05:50PM +0000, Titus Rwantare wrote: > PECI - Platform Environment Control Interface > > This commit adds support for reading basic sensor values from a client > on the PECI bus. > BMCs can use the PECI wire to get thermal information out of an Intel > cpu. Additionally, on hardware, various MSRs are exposed over the > PECI bus. Part of PCI config space is exposed due to Intel posting > various platform configuration in PCI config space. > > Commands implemented: > - Ping > - GetDIB > - GetTemp > - GetPkgConfig (partial) > > Commands not implemented: > - RdIAMSR > - RdPCIConfig > - RdPCIConfigLocal > > Signed-off-by: Titus Rwantare <titusr@google.com> > Reviewed-by: Hao Wu <wuhaotsh@google.com> > Reviewed-by: Patrick Venture <venture@google.com> > --- > MAINTAINERS | 7 ++ > hw/Kconfig | 1 + > hw/meson.build | 1 + > hw/peci/Kconfig | 2 + > hw/peci/meson.build | 1 + > hw/peci/peci-client.c | 230 +++++++++++++++++++++++++++++++++++++++++ > hw/peci/peci-core.c | 182 ++++++++++++++++++++++++++++++++ > hw/peci/trace-events | 5 + > hw/peci/trace.h | 1 + > include/hw/peci/peci.h | 194 ++++++++++++++++++++++++++++++++++ > meson.build | 1 + > 11 files changed, 625 insertions(+) > create mode 100644 hw/peci/Kconfig > create mode 100644 hw/peci/meson.build > create mode 100644 hw/peci/peci-client.c > create mode 100644 hw/peci/peci-core.c > create mode 100644 hw/peci/trace-events > create mode 100644 hw/peci/trace.h > create mode 100644 include/hw/peci/peci.h > > diff --git a/MAINTAINERS b/MAINTAINERS > index 5ce4227ff6..14ab29679d 100644 > --- a/MAINTAINERS > +++ b/MAINTAINERS > @@ -3212,6 +3212,13 @@ F: tests/qtest/adm1272-test.c > F: tests/qtest/max34451-test.c > F: tests/qtest/isl_pmbus_vr-test.c > > +PECI > +M: Titus Rwantare <titusr@google.com> > +S: Maintained > +F: hw/peci/peci-core.c > +F: hw/peci/peci-client.c > +F: include/hw/peci/peci.h > + > Firmware schema specifications > M: Philippe Mathieu-Daudé <f4bug@amsat.org> > R: Daniel P. Berrange <berrange@redhat.com> > diff --git a/hw/Kconfig b/hw/Kconfig > index 38233bbb0f..300ab48127 100644 > --- a/hw/Kconfig > +++ b/hw/Kconfig > @@ -24,6 +24,7 @@ source net/Kconfig > source nubus/Kconfig > source nvme/Kconfig > source nvram/Kconfig > +source peci/Kconfig > source pci-bridge/Kconfig > source pci-host/Kconfig > source pcmcia/Kconfig > diff --git a/hw/meson.build b/hw/meson.build > index c7ac7d3d75..340cc88a52 100644 > --- a/hw/meson.build > +++ b/hw/meson.build > @@ -28,6 +28,7 @@ subdir('pci') > subdir('pci-bridge') > subdir('pci-host') > subdir('pcmcia') > +subdir('peci') > subdir('rdma') > subdir('rtc') > subdir('scsi') > diff --git a/hw/peci/Kconfig b/hw/peci/Kconfig > new file mode 100644 > index 0000000000..fe4f665d21 > --- /dev/null > +++ b/hw/peci/Kconfig > @@ -0,0 +1,2 @@ > +config PECI > + bool > diff --git a/hw/peci/meson.build b/hw/peci/meson.build > new file mode 100644 > index 0000000000..01cfa95abe > --- /dev/null > +++ b/hw/peci/meson.build > @@ -0,0 +1 @@ > +softmmu_ss.add(when: 'CONFIG_PECI', if_true: files('peci-core.c', 'peci-client.c')) > diff --git a/hw/peci/peci-client.c b/hw/peci/peci-client.c > new file mode 100644 > index 0000000000..2aa797b5f6 > --- /dev/null > +++ b/hw/peci/peci-client.c > @@ -0,0 +1,230 @@ > +/* > + * PECI Client device > + * Copyright 2021 Google LLC > + * > + * SPDX-License-Identifier: GPL-2.0-or-later Not sure, but I think the SPDX license identifier is supposed to be in the first line? Maybe not though. I would have expected: /* SPDX-License-Identifier: GPL-2.0-or-later */ /* * PECI Client device * * Copyright 2021 Google LLC */ > + */ > + > +#include "qemu/osdep.h" > +#include "hw/peci/peci.h" > +#include "hw/qdev-core.h" > +#include "hw/qdev-properties.h" > +#include "migration/vmstate.h" > +#include "qapi/error.h" > +#include "qapi/visitor.h" > +#include "qemu/module.h" > +#include "qemu/log.h" > +#include "qom/object.h" > + > +/* > + * A PECI client represents an Intel socket and the peripherals attached to it > + * that are accessible over the PECI bus. > + */ > + > +#define PECI_CLIENT_DEFAULT_TEMP 30 > + > +static void peci_client_update_temps(PECIClientDevice *client) > +{ > + uint8_t temp_cpu = 0; > + for (size_t i = 0; i < client->cpu_cores; i++) { > + if (temp_cpu < client->core_temp[i]) { > + temp_cpu = client->core_temp[i]; > + } > + } > + client->core_temp_max = -1 * (client->tjmax - temp_cpu); > + > + uint8_t temp_dimm = 0; > + for (size_t i = 0; i < client->dimms; i++) { > + if (temp_dimm < client->dimm_temp[i]) { > + temp_dimm = client->dimm_temp[i]; > + } > + } > + client->dimm_temp_max = temp_dimm; > +} > + > +PECIClientDevice *peci_get_client(PECIBus *bus, uint8_t addr) > +{ > + PECIClientDevice *client; > + BusChild *var; > + > + QTAILQ_FOREACH(var, &bus->qbus.children, sibling) { > + DeviceState *dev = var->child; > + client = PECI_CLIENT(dev); > + > + if (client->address == addr) { > + return client; > + } > + } > + return 0; > +} > + > + > +PECIClientDevice *peci_add_client(PECIBus *bus, > + uint8_t address, > + PECIClientProperties *props) > +{ > + DeviceState *dev = qdev_new("peci-client"); > + PECIClientDevice *client; > + > + /* Only 8 addresses supported as of rev 4.1, 0x30 to 0x37 */ > + if (address < PECI_BASE_ADDR || address > PECI_BASE_ADDR + 7) { > + qemu_log_mask(LOG_GUEST_ERROR, "%s: failed to add client at 0x%02x", > + __func__, address); > + return 0; > + } > + > + client = PECI_CLIENT(dev); > + client->address = address; > + > + /* these fields are optional, get set to max if unset or invalid */ > + if (!props->cpu_cores || props->cpu_cores > PECI_NUM_CPUS_MAX) { > + props->cpu_cores = PECI_NUM_CPUS_MAX; > + } > + if (!props->dimms || props->dimms > PECI_NUM_DIMMS_MAX) { > + props->dimms = PECI_NUM_DIMMS_MAX; > + } > + if (!props->cpu_family) { > + props->cpu_family = FAM6_ICELAKE_X; > + } > + if (!props->dimms_per_channel || > + props->dimms_per_channel > PECI_DIMMS_PER_CHANNEL_MAX) { > + props->dimms_per_channel = 2; > + } > + > + client->cpu_id = props->cpu_family; > + client->cpu_cores = props->cpu_cores; > + client->dimms = props->dimms; > + client->dimms_per_channel = props->dimms_per_channel; > + > + /* TODO: find real revisions and TJ max for supported families */ > + client->tjmax = 90; > + client->tcontrol = -5; > + > + switch (props->cpu_family) { > + case FAM6_HASWELL_X: > + client->revision = 0x31; > + break; > + > + case FAM6_BROADWELL_X: > + client->revision = 0x32; > + break; > + > + case FAM6_SKYLAKE_X: > + case FAM6_SKYLAKE_XD: > + client->revision = 0x36; > + break; > + > + case FAM6_ICELAKE_X: > + case FAM6_SAPPHIRE_RAPIDS_X: > + client->revision = 0x40; > + client->ucode = 0x8c0004a0; > + break; > + > + default: > + qemu_log_mask(LOG_UNIMP, "%s: unsupported cpu: 0x%x\n", > + __func__, props->cpu_family); > + client->revision = 0x31; > + break; > + } > + > + qdev_realize_and_unref(&client->qdev, &bus->qbus, &error_abort); > + bus->num_clients += 1; > + return client; > +} > + > +static void peci_client_get(Object *obj, Visitor *v, const char *name, > + void *opaque, Error **errp) > +{ > + /* use millidegrees convention */ > + uint32_t value = *(uint8_t *)opaque * 1000; > + visit_type_uint32(v, name, &value, errp); > +} > + > +static void peci_client_set(Object *obj, Visitor *v, const char *name, > + void *opaque, Error **errp) > +{ > + PECIClientDevice *client = PECI_CLIENT(obj); > + uint8_t *internal = opaque; > + uint32_t value; > + > + if (!visit_type_uint32(v, name, &value, errp)) { > + return; > + } > + > + g_assert(value <= 255000); > + > + *internal = value / 1000; > + peci_client_update_temps(client); > +} > + > +static void peci_client_reset(Object *obj, ResetType type) > +{ > + PECIClientDevice *client = PECI_CLIENT(obj); > + client->core_temp_max = 0; > + client->dimm_temp_max = 0; > +} > + > +static const VMStateDescription vmstate_peci_client = { > + .name = TYPE_PECI_CLIENT, > + .version_id = 0, > + .minimum_version_id = 0, > + .fields = (VMStateField[]) { > + VMSTATE_UINT8(address, PECIClientDevice), > + VMSTATE_UINT8(device_info, PECIClientDevice), > + VMSTATE_UINT8(revision, PECIClientDevice), > + VMSTATE_UINT32(cpu_id, PECIClientDevice), > + VMSTATE_UINT8(cpu_cores, PECIClientDevice), > + VMSTATE_UINT8(tjmax, PECIClientDevice), > + VMSTATE_INT8(tcontrol, PECIClientDevice), > + VMSTATE_UINT8_ARRAY(core_temp, PECIClientDevice, PECI_NUM_CPUS_MAX), > + VMSTATE_UINT8(dimms, PECIClientDevice), > + VMSTATE_UINT8(dimms_per_channel, PECIClientDevice), > + VMSTATE_UINT8_ARRAY(dimm_temp, PECIClientDevice , PECI_NUM_DIMMS_MAX), > + VMSTATE_END_OF_LIST() > + }, > +}; > + > +static void peci_client_realize(DeviceState *dev, Error **errp) > +{ > + PECIClientDevice *client = PECI_CLIENT(dev); > + for (size_t i = 0; i < client->cpu_cores; i++) { > + client->core_temp[i] = PECI_CLIENT_DEFAULT_TEMP; > + object_property_add(OBJECT(client), "cpu_temp[*]", "uint8", > + peci_client_get, > + peci_client_set, NULL, &client->core_temp[i]); > + } > + > + for (size_t i = 0; i < client->dimms; i++) { > + client->dimm_temp[i] = PECI_CLIENT_DEFAULT_TEMP; > + object_property_add(OBJECT(client), "dimm_temp[*]", "uint8", > + peci_client_get, > + peci_client_set, NULL, &client->dimm_temp[i]); > + } > +} > + > +static void peci_client_class_init(ObjectClass *klass, void *data) > +{ > + DeviceClass *dc = DEVICE_CLASS(klass); > + ResettableClass *rc = RESETTABLE_CLASS(klass); > + > + dc->desc = "PECI Client Device"; > + dc->bus_type = TYPE_PECI_BUS; > + dc->realize = peci_client_realize; > + dc->vmsd = &vmstate_peci_client; > + set_bit(DEVICE_CATEGORY_MISC, dc->categories); > + rc->phases.enter = peci_client_reset; > +} > + > +static const TypeInfo peci_client_info = { > + .name = TYPE_PECI_CLIENT, > + .parent = TYPE_DEVICE, > + .instance_size = sizeof(PECIClientDevice), > + .class_init = peci_client_class_init, > +}; > + > +static void peci_client_register_types(void) > +{ > + type_register_static(&peci_client_info); > +} > + > +type_init(peci_client_register_types) > diff --git a/hw/peci/peci-core.c b/hw/peci/peci-core.c > new file mode 100644 > index 0000000000..8210bfa198 > --- /dev/null > +++ b/hw/peci/peci-core.c > @@ -0,0 +1,182 @@ > +/* > + * PECI - Platform Environment Control Interface > + * Copyright 2021 Google LLC > + * > + * SPDX-License-Identifier: GPL-2.0-or-later > + */ > + > +/* > + * This PECI implementation aims to simulate a host with peci as viewed by a > + * BMC. This is developed with OpenBMC firmware running in QEMU. > + * > + */ > + > +#include "qemu/osdep.h" > +#include "hw/peci/peci.h" > +#include "hw/qdev-properties.h" > +#include "migration/vmstate.h" > +#include "qemu/module.h" > +#include "qemu/log.h" > +#include "trace.h" > + > +#define PECI_FCS_OK 0 > +#define PECI_FCS_ERR 1 > + > +static void peci_rd_pkg_cfg(PECIClientDevice *client, PECICmd *pcmd) > +{ > + PECIPkgCfg *resp = (PECIPkgCfg *)pcmd->tx; > + uint8_t index = pcmd->rx[1]; > + uint16_t param = pcmd->rx[3] | pcmd->rx[2]; > + > + switch (index) { > + case PECI_MBX_CPU_ID: /* CPU Family ID*/ > + trace_peci_rd_pkg_cfg("CPU ID"); > + switch (param) { > + case PECI_PKG_ID_CPU_ID: > + resp->data = client->cpu_id; > + break; > + > + case PECI_PKG_ID_MICROCODE_REV: > + resp->data = client->ucode; > + break; > + > + default: > + qemu_log_mask(LOG_UNIMP, "%s: CPU ID param %u\n", __func__, param); > + break; > + } > + break; > + > + case PECI_MBX_DTS_MARGIN: > + trace_peci_rd_pkg_cfg("DTS MARGIN"); > + /* > + * Processors return a value of DTS reading in 10.6 format > + * (10 bits signed decimal, 6 bits fractional). > + */ > + resp->data = (-1 * client->core_temp_max) << 6; > + break; > + > + case PECI_MBX_DDR_DIMM_TEMP: > + trace_peci_rd_pkg_cfg("DDR DIMM TEMP"); > + if (param > PECI_NUM_CHANNELS_MAX) { > + qemu_log_mask(LOG_GUEST_ERROR, > + "%s: DDR_DIMM_TEMP unsupported channel index %d", > + __func__, param); > + param = PECI_NUM_CHANNELS_MAX; > + } > + uint8_t channel_index = param * client->dimms_per_channel; > + memcpy(&resp->data, > + &client->dimm_temp[channel_index], > + sizeof(client->dimm_temp[0]) * client->dimms_per_channel); > + break; > + > + case PECI_MBX_TEMP_TARGET: > + trace_peci_rd_pkg_cfg("TEMP TARGET"); > + PECITempTarget target = { > + .tcontrol = client->tcontrol, > + .tjmax = client->tjmax, > + }; > + memcpy(&resp->data, &target, sizeof(target)); > + break; > + > + case PECI_MBX_PPIN: > + trace_peci_rd_pkg_cfg("PPIN"); > + resp->data = 0xdeadbeef; > + break; > + > + default: > + qemu_log_mask(LOG_UNIMP, "%s: unimplemented PkgCfg Index: %u\n", > + __func__, index); > + resp->cc = PECI_DEV_CC_INVALID_REQ; > + return; > + } > + > + resp->cc = PECI_DEV_CC_SUCCESS; > +} > + > +int peci_handle_cmd(PECIBus *bus, PECICmd *pcmd) > +{ > + PECIClientDevice *client = peci_get_client(bus, pcmd->addr); > + > + if (!client) { /* ignore commands if clients aren't found */ > + qemu_log_mask(LOG_GUEST_ERROR, > + "%s: no peci client at address: 0x%02x\n", > + __func__, pcmd->addr); > + return PECI_FCS_ERR; > + } > + > + /* clear output buffer before handling command */ > + for (size_t i = 0; i < pcmd->rd_length; i++) { > + pcmd->tx[i] = 0; > + } > + > + switch (pcmd->cmd) { > + case PECI_CMD_PING: > + trace_peci_handle_cmd("PING!", pcmd->addr); > + break; > + > + case PECI_CMD_GET_DIB: /* Device Info Byte */ > + trace_peci_handle_cmd("GetDIB", pcmd->addr); > + PECIDIB dib = { > + .device_info = client->device_info, > + .revision = client->revision, > + }; > + memcpy(pcmd->tx, &dib, sizeof(PECIDIB)); > + break; > + > + case PECI_CMD_GET_TEMP: /* maximum die temp in socket */ > + trace_peci_handle_cmd("GetTemp", pcmd->addr); > + /* > + * The data is returned as a negative value representing the number of > + * degrees centigrade below the maximum processor junction temperature > + */ > + memcpy(pcmd->tx, &client->core_temp_max, sizeof(client->core_temp_max)); > + break; > + > + case PECI_CMD_RD_PKG_CFG: > + trace_peci_handle_cmd("RdPkgConfig", pcmd->addr); > + peci_rd_pkg_cfg(client, pcmd); > + break; > + > + case PECI_CMD_RD_IA_MSR: > + trace_peci_handle_cmd("RdIAMSR", pcmd->addr); > + qemu_log_mask(LOG_UNIMP, "%s: unimplemented RD_IA_MSR\n", __func__); > + break; > + > + case PECI_CMD_RD_PCI_CFG: > + trace_peci_handle_cmd("RdPCIConfig", pcmd->addr); > + qemu_log_mask(LOG_UNIMP, "%s: unimplemented PCI_CFG\n", __func__); > + break; > + > + case PECI_CMD_RD_PCI_CFG_LOCAL: > + trace_peci_handle_cmd("RdPCIConfigLocal", pcmd->addr); > + qemu_log_mask(LOG_UNIMP, "%s: unimplemented PCI_CFG_LOCAL\n", __func__); > + break; > + > + case PECI_CMD_RD_END_PT_CFG: > + qemu_log_mask(LOG_UNIMP, "%s: unimplemented CMD_RD_END_PT_CFG\n", > + __func__); > + break; > + > + default: > + qemu_log_mask(LOG_GUEST_ERROR, "%s: unknown cmd: %x\n", > + __func__, pcmd->cmd); > + break; > + } > + > + return PECI_FCS_OK; > +} > + > +PECIBus *peci_bus_create(DeviceState *parent) > +{ > + return PECI_BUS(qbus_new(TYPE_PECI_BUS, parent, "peci-bus")); > +}; > + > +static const TypeInfo peci_types[] = { > + { > + .name = TYPE_PECI_BUS, > + .parent = TYPE_BUS, > + .instance_size = sizeof(PECIBus), > + }, > +}; > + > +DEFINE_TYPES(peci_types); > diff --git a/hw/peci/trace-events b/hw/peci/trace-events > new file mode 100644 > index 0000000000..f90c998dd9 > --- /dev/null > +++ b/hw/peci/trace-events > @@ -0,0 +1,5 @@ > +# See docs/devel/tracing.rst for syntax documentation. > + > +# peci-core.c > +peci_handle_cmd(const char* s, uint8_t addr) "%s @ 0x%02" PRIx8 > +peci_rd_pkg_cfg(const char* s) "%s" > diff --git a/hw/peci/trace.h b/hw/peci/trace.h > new file mode 100644 > index 0000000000..4283e3bfc5 > --- /dev/null > +++ b/hw/peci/trace.h > @@ -0,0 +1 @@ > +#include "trace/trace-hw_peci.h" > diff --git a/include/hw/peci/peci.h b/include/hw/peci/peci.h > new file mode 100644 > index 0000000000..1a0abe65cd > --- /dev/null > +++ b/include/hw/peci/peci.h > @@ -0,0 +1,194 @@ > +/* > + * PECI - Platform Environment Control Interface > + * > + * Copyright 2021 Google LLC > + * > + * SPDX-License-Identifier: GPL-2.0-or-later > + */ > + > +#ifndef PECI_H > +#define PECI_H > + > +#include "hw/qdev-core.h" > +#include "qom/object.h" > + > +#define PECI_CMD_PING 0x0 > +#define PECI_CMD_GET_DIB 0xF7 > +#define PECI_CMD_GET_TEMP 0x01 > +#define PECI_CMD_RD_PKG_CFG 0xA1 > +#define PECI_CMD_WR_PKG_CFG 0xA5 > +#define PECI_MBX_CPU_ID 0 /* Package Identifier Read */ > +#define PECI_PKG_ID_CPU_ID 0x0000 /* CPUID Info */ > +#define PECI_PKG_ID_PLATFORM_ID 0x0001 /* Platform ID */ > +#define PECI_PKG_ID_UNCORE_ID 0x0002 /* Uncore Device ID */ > +#define PECI_PKG_ID_MAX_THREAD_ID 0x0003 /* Max Thread ID */ > +#define PECI_PKG_ID_MICROCODE_REV 0x0004 /* CPU Microcode Revision */ > +#define PECI_PKG_ID_MACHINE_CHECK_STATUS 0x0005 /* Machine Check Status */ > +#define PECI_MBX_VR_DEBUG 1 /* VR Debug */ > +#define PECI_MBX_PKG_TEMP_READ 2 /* Package Temperature Read */ > +#define PECI_MBX_ENERGY_COUNTER 3 /* Energy counter */ > +#define PECI_MBX_ENERGY_STATUS 4 /* DDR Energy Status */ > +#define PECI_MBX_WAKE_MODE_BIT 5 /* "Wake on PECI" Mode bit */ > +#define PECI_MBX_EPI 6 /* Efficient Performance Indication */ > +#define PECI_MBX_PKG_RAPL_PERF 8 /* Pkg RAPL Performance Status Read */ > +#define PECI_MBX_PER_CORE_DTS_TEMP 9 /* Per Core DTS Temperature Read */ > +#define PECI_MBX_DTS_MARGIN 10 /* DTS thermal margin */ > +#define PECI_MBX_SKT_PWR_THRTL_DUR 11 /* Socket Power Throttled Duration */ > +#define PECI_MBX_CFG_TDP_CONTROL 12 /* TDP Config Control */ > +#define PECI_MBX_CFG_TDP_LEVELS 13 /* TDP Config Levels */ > +#define PECI_MBX_DDR_DIMM_TEMP 14 /* DDR DIMM Temperature */ > +#define PECI_MBX_CFG_ICCMAX 15 /* Configurable ICCMAX */ > +#define PECI_MBX_TEMP_TARGET 16 /* Temperature Target Read */ > +#define PECI_MBX_CURR_CFG_LIMIT 17 /* Current Config Limit */ > +#define PECI_MBX_PPIN 19 /* Processor Inventory Number */ > +#define PECI_MBX_DIMM_TEMP_READ 20 /* Package Thermal Status Read */ > +#define PECI_MBX_DRAM_IMC_TMP_READ 22 /* DRAM IMC Temperature Read */ > +#define PECI_MBX_DDR_CH_THERM_STAT 23 /* DDR Channel Thermal Status */ > +#define PECI_MBX_PKG_POWER_LIMIT1 26 /* Package Power Limit1 */ > +#define PECI_MBX_PKG_POWER_LIMIT2 27 /* Package Power Limit2 */ > +#define PECI_MBX_TDP 28 /* Thermal design power minimum */ > +#define PECI_MBX_TDP_HIGH 29 /* Thermal design power maximum */ > +#define PECI_MBX_TDP_UNITS 30 /* Units for power/energy registers */ > +#define PECI_MBX_RUN_TIME 31 /* Accumulated Run Time */ > +#define PECI_MBX_CONSTRAINED_TIME 32 /* Thermally Constrained Time Read */ > +#define PECI_MBX_TURBO_RATIO 33 /* Turbo Activation Ratio */ > +#define PECI_MBX_DDR_RAPL_PL1 34 /* DDR RAPL PL1 */ > +#define PECI_MBX_DDR_PWR_INFO_HIGH 35 /* DRAM Power Info Read (high) */ > +#define PECI_MBX_DDR_PWR_INFO_LOW 36 /* DRAM Power Info Read (low) */ > +#define PECI_MBX_DDR_RAPL_PL2 37 /* DDR RAPL PL2 */ > +#define PECI_MBX_DDR_RAPL_STATUS 38 /* DDR RAPL Performance Status */ > +#define PECI_MBX_DDR_HOT_ABSOLUTE 43 /* DDR Hottest Dimm Absolute Temp */ > +#define PECI_MBX_DDR_HOT_RELATIVE 44 /* DDR Hottest Dimm Relative Temp */ > +#define PECI_MBX_DDR_THROTTLE_TIME 45 /* DDR Throttle Time */ > +#define PECI_MBX_DDR_THERM_STATUS 46 /* DDR Thermal Status */ > +#define PECI_MBX_TIME_AVG_TEMP 47 /* Package time-averaged temperature */ > +#define PECI_MBX_TURBO_RATIO_LIMIT 49 /* Turbo Ratio Limit Read */ > +#define PECI_MBX_HWP_AUTO_OOB 53 /* HWP Autonomous Out-of-band */ > +#define PECI_MBX_DDR_WARM_BUDGET 55 /* DDR Warm Power Budget */ > +#define PECI_MBX_DDR_HOT_BUDGET 56 /* DDR Hot Power Budget */ > +#define PECI_MBX_PKG_PSYS_PWR_LIM3 57 /* Package/Psys Power Limit3 */ > +#define PECI_MBX_PKG_PSYS_PWR_LIM1 58 /* Package/Psys Power Limit1 */ > +#define PECI_MBX_PKG_PSYS_PWR_LIM2 59 /* Package/Psys Power Limit2 */ > +#define PECI_MBX_PKG_PSYS_PWR_LIM4 60 /* Package/Psys Power Limit4 */ > +#define PECI_MBX_PERF_LIMIT_REASON 65 /* Performance Limit Reasons */ > +#define PECI_CMD_RD_IA_MSR 0xB1 > +#define PECI_CMD_WR_IA_MSR 0xB5 > +#define PECI_CMD_RD_IA_MSREX 0xD1 > +#define PECI_CMD_RD_PCI_CFG 0x61 > +#define PECI_CMD_WR_PCI_CFG 0x65 > +#define PECI_CMD_RD_PCI_CFG_LOCAL 0xE1 > +#define PECI_CMD_WR_PCI_CFG_LOCAL 0xE5 > +#define PECI_CMD_RD_END_PT_CFG 0xC1 > +#define PECI_CMD_WR_END_PT_CFG 0xC5 > +#define PECI_CMD_CRASHDUMP_GET_FRAME 0x71 > + > +/* Device Specific Completion Code (CC) Definition */ > +#define PECI_DEV_CC_SUCCESS 0x40 > +#define PECI_DEV_CC_NEED_RETRY 0x80 > +#define PECI_DEV_CC_OUT_OF_RESOURCE 0x81 > +#define PECI_DEV_CC_UNAVAIL_RESOURCE 0x82 > +#define PECI_DEV_CC_INVALID_REQ 0x90 > +#define PECI_DEV_CC_MCA_ERROR 0x91 > +#define PECI_DEV_CC_CATASTROPHIC_MCA_ERROR 0x93 > +#define PECI_DEV_CC_FATAL_MCA_DETECTED 0x94 > +#define PECI_DEV_CC_PARITY_ERROR_ON_GPSB_OR_PMSB 0x98 > +#define PECI_DEV_CC_PARITY_ERROR_ON_GPSB_OR_PMSB_IERR 0x9B > +#define PECI_DEV_CC_PARITY_ERROR_ON_GPSB_OR_PMSB_MCA 0x9C > + > + > +typedef struct PECIDIB { /* DIB - Device Info Byte(s) */ > + uint8_t device_info; /* bit 2 - num of domains */ > + uint8_t revision; > +} PECIDIB; > + > +typedef struct __attribute__ ((__packed__)) { > + uint8_t cc; /* completion code */ > + uint32_t data; > +} PECIPkgCfg; > + > +typedef struct PECITempTarget { > + uint8_t reserved; > + int8_t tcontrol; > + uint8_t tjmax; > +} PECITempTarget; > + > +#define PECI_BASE_ADDR 0x30 > +#define PECI_BUFFER_SIZE 0x100 > +#define PECI_NUM_CPUS_MAX 56 > +#define PECI_DIMMS_PER_CHANNEL_MAX 3 > +#define PECI_NUM_CHANNELS_MAX 8 > +#define PECI_NUM_DIMMS_MAX (PECI_NUM_CHANNELS_MAX * PECI_DIMMS_PER_CHANNEL_MAX) > + > +typedef struct PECIClientDevice { > + DeviceState qdev; > + uint8_t address; > + uint8_t device_info; > + uint8_t revision; > + > + /* CPU */ > + uint32_t cpu_id; > + uint8_t cpu_cores; > + uint32_t ucode; > + uint8_t tjmax; > + int8_t tcontrol; > + int8_t core_temp_max; /* absolute temp - tjmax */ > + uint8_t core_temp[PECI_NUM_CPUS_MAX]; > + > + /* DIMMS */ > + uint8_t dimms; > + uint8_t dimms_per_channel; > + uint8_t dimm_temp_max; > + uint8_t dimm_temp[PECI_NUM_DIMMS_MAX]; > + > +} PECIClientDevice; > + > +#define TYPE_PECI_CLIENT "peci-client" > +OBJECT_DECLARE_SIMPLE_TYPE(PECIClientDevice, PECI_CLIENT) > + > +typedef struct PECIBus { > + BusState qbus; > + uint8_t num_clients; > +} PECIBus; > + > +#define TYPE_PECI_BUS "peci-bus" > +OBJECT_DECLARE_SIMPLE_TYPE(PECIBus, PECI_BUS) > + > +/* Creates a basic qemu bus onto which PECI clients will be attached */ > +PECIBus *peci_bus_create(DeviceState *parent); > + > +enum { > + FAM6_HASWELL_X = 0x306F0, > + FAM6_BROADWELL_X = 0x406F0, > + FAM6_SKYLAKE_X = 0x50650, > + FAM6_SKYLAKE_XD = 0x50660, > + FAM6_ICELAKE_X = 0x606A0, > + FAM6_SAPPHIRE_RAPIDS_X = 0x806F3, > +}; I'm curious if we really need the CPU family here, or if we could just base everything off the PECI version. The PECI specification doesn't mention the CPU family, does it? Or maybe it does. > + > +typedef struct PECIClientProperties { > + uint32_t cpu_family; > + uint8_t cpu_cores; > + uint8_t dimms; > + uint8_t dimms_per_channel; > +} PECIClientProperties; > + > +/* Creates a PECI client with the specified cpu and dimm count */ > +PECIClientDevice *peci_add_client(PECIBus *bus, > + uint8_t address, > + PECIClientProperties *props); > +PECIClientDevice *peci_get_client(PECIBus *bus, uint8_t addr); > + > +typedef struct PECICmd { > + uint8_t addr; > + uint8_t cmd; > + > + size_t wr_length; > + uint8_t rx[PECI_BUFFER_SIZE]; > + > + size_t rd_length; > + uint8_t tx[PECI_BUFFER_SIZE]; > +} PECICmd; > + > +int peci_handle_cmd(PECIBus *bus, PECICmd *pcmd); > + > +#endif > diff --git a/meson.build b/meson.build > index 20fddbd707..55594702c3 100644 > --- a/meson.build > +++ b/meson.build > @@ -2965,6 +2965,7 @@ if have_system > 'hw/nvram', > 'hw/pci', > 'hw/pci-host', > + 'hw/peci', > 'hw/ppc', > 'hw/rdma', > 'hw/rdma/vmw', > -- > 2.37.2.789.g6183377224-goog >
On Fri, 9 Sept 2022 at 12:58, Peter Delevoryas <peter@pjd.dev> wrote: > > +/* > > + * PECI Client device > > + * Copyright 2021 Google LLC > > + * > > + * SPDX-License-Identifier: GPL-2.0-or-later > > Not sure, but I think the SPDX license identifier is supposed to be in > the first line? Maybe not though. I would have expected: > That's a Linux thing as far as I can tell. QEMU has it in the top comment. > > I'm curious if we really need the CPU family here, or if we could just > base everything off the PECI version. > > The PECI specification doesn't mention the CPU family, does it? Or maybe > it does. > I needed the family info anyway for RdPkgConfig() CPU ID, and thought it would be more readable to specify that in the board file than the PECI version. We tend to add new machines to QEMU by copying the config of an existing machine, I think this way makes it more obvious that this is a field that is changing. Titus
On Tue, Sep 13, 2022 at 11:21:16AM -0700, Titus Rwantare wrote: > On Fri, 9 Sept 2022 at 12:58, Peter Delevoryas <peter@pjd.dev> wrote: > > > > +/* > > > + * PECI Client device > > > + * Copyright 2021 Google LLC > > > + * > > > + * SPDX-License-Identifier: GPL-2.0-or-later > > > > Not sure, but I think the SPDX license identifier is supposed to be in > > the first line? Maybe not though. I would have expected: > > > > That's a Linux thing as far as I can tell. QEMU has it in the top comment. Oh ok, nevermind then. > > > > > I'm curious if we really need the CPU family here, or if we could just > > base everything off the PECI version. > > > > The PECI specification doesn't mention the CPU family, does it? Or maybe > > it does. > > > > I needed the family info anyway for RdPkgConfig() CPU ID, and thought it > would be more readable to specify that in the board file than the PECI version. > We tend to add new machines to QEMU by copying the config of an existing > machine, I think this way makes it more obvious that this is a field > that is changing. Yeah, I think you're right. And looking at the PECI spec, they refer to families directly when describing new features, so this seems appropriate. Reviewed-by: Peter Delevoryas <peter@pjd.dev> > > > Titus
diff --git a/MAINTAINERS b/MAINTAINERS index 5ce4227ff6..14ab29679d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3212,6 +3212,13 @@ F: tests/qtest/adm1272-test.c F: tests/qtest/max34451-test.c F: tests/qtest/isl_pmbus_vr-test.c +PECI +M: Titus Rwantare <titusr@google.com> +S: Maintained +F: hw/peci/peci-core.c +F: hw/peci/peci-client.c +F: include/hw/peci/peci.h + Firmware schema specifications M: Philippe Mathieu-Daudé <f4bug@amsat.org> R: Daniel P. Berrange <berrange@redhat.com> diff --git a/hw/Kconfig b/hw/Kconfig index 38233bbb0f..300ab48127 100644 --- a/hw/Kconfig +++ b/hw/Kconfig @@ -24,6 +24,7 @@ source net/Kconfig source nubus/Kconfig source nvme/Kconfig source nvram/Kconfig +source peci/Kconfig source pci-bridge/Kconfig source pci-host/Kconfig source pcmcia/Kconfig diff --git a/hw/meson.build b/hw/meson.build index c7ac7d3d75..340cc88a52 100644 --- a/hw/meson.build +++ b/hw/meson.build @@ -28,6 +28,7 @@ subdir('pci') subdir('pci-bridge') subdir('pci-host') subdir('pcmcia') +subdir('peci') subdir('rdma') subdir('rtc') subdir('scsi') diff --git a/hw/peci/Kconfig b/hw/peci/Kconfig new file mode 100644 index 0000000000..fe4f665d21 --- /dev/null +++ b/hw/peci/Kconfig @@ -0,0 +1,2 @@ +config PECI + bool diff --git a/hw/peci/meson.build b/hw/peci/meson.build new file mode 100644 index 0000000000..01cfa95abe --- /dev/null +++ b/hw/peci/meson.build @@ -0,0 +1 @@ +softmmu_ss.add(when: 'CONFIG_PECI', if_true: files('peci-core.c', 'peci-client.c')) diff --git a/hw/peci/peci-client.c b/hw/peci/peci-client.c new file mode 100644 index 0000000000..2aa797b5f6 --- /dev/null +++ b/hw/peci/peci-client.c @@ -0,0 +1,230 @@ +/* + * PECI Client device + * Copyright 2021 Google LLC + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "hw/peci/peci.h" +#include "hw/qdev-core.h" +#include "hw/qdev-properties.h" +#include "migration/vmstate.h" +#include "qapi/error.h" +#include "qapi/visitor.h" +#include "qemu/module.h" +#include "qemu/log.h" +#include "qom/object.h" + +/* + * A PECI client represents an Intel socket and the peripherals attached to it + * that are accessible over the PECI bus. + */ + +#define PECI_CLIENT_DEFAULT_TEMP 30 + +static void peci_client_update_temps(PECIClientDevice *client) +{ + uint8_t temp_cpu = 0; + for (size_t i = 0; i < client->cpu_cores; i++) { + if (temp_cpu < client->core_temp[i]) { + temp_cpu = client->core_temp[i]; + } + } + client->core_temp_max = -1 * (client->tjmax - temp_cpu); + + uint8_t temp_dimm = 0; + for (size_t i = 0; i < client->dimms; i++) { + if (temp_dimm < client->dimm_temp[i]) { + temp_dimm = client->dimm_temp[i]; + } + } + client->dimm_temp_max = temp_dimm; +} + +PECIClientDevice *peci_get_client(PECIBus *bus, uint8_t addr) +{ + PECIClientDevice *client; + BusChild *var; + + QTAILQ_FOREACH(var, &bus->qbus.children, sibling) { + DeviceState *dev = var->child; + client = PECI_CLIENT(dev); + + if (client->address == addr) { + return client; + } + } + return 0; +} + + +PECIClientDevice *peci_add_client(PECIBus *bus, + uint8_t address, + PECIClientProperties *props) +{ + DeviceState *dev = qdev_new("peci-client"); + PECIClientDevice *client; + + /* Only 8 addresses supported as of rev 4.1, 0x30 to 0x37 */ + if (address < PECI_BASE_ADDR || address > PECI_BASE_ADDR + 7) { + qemu_log_mask(LOG_GUEST_ERROR, "%s: failed to add client at 0x%02x", + __func__, address); + return 0; + } + + client = PECI_CLIENT(dev); + client->address = address; + + /* these fields are optional, get set to max if unset or invalid */ + if (!props->cpu_cores || props->cpu_cores > PECI_NUM_CPUS_MAX) { + props->cpu_cores = PECI_NUM_CPUS_MAX; + } + if (!props->dimms || props->dimms > PECI_NUM_DIMMS_MAX) { + props->dimms = PECI_NUM_DIMMS_MAX; + } + if (!props->cpu_family) { + props->cpu_family = FAM6_ICELAKE_X; + } + if (!props->dimms_per_channel || + props->dimms_per_channel > PECI_DIMMS_PER_CHANNEL_MAX) { + props->dimms_per_channel = 2; + } + + client->cpu_id = props->cpu_family; + client->cpu_cores = props->cpu_cores; + client->dimms = props->dimms; + client->dimms_per_channel = props->dimms_per_channel; + + /* TODO: find real revisions and TJ max for supported families */ + client->tjmax = 90; + client->tcontrol = -5; + + switch (props->cpu_family) { + case FAM6_HASWELL_X: + client->revision = 0x31; + break; + + case FAM6_BROADWELL_X: + client->revision = 0x32; + break; + + case FAM6_SKYLAKE_X: + case FAM6_SKYLAKE_XD: + client->revision = 0x36; + break; + + case FAM6_ICELAKE_X: + case FAM6_SAPPHIRE_RAPIDS_X: + client->revision = 0x40; + client->ucode = 0x8c0004a0; + break; + + default: + qemu_log_mask(LOG_UNIMP, "%s: unsupported cpu: 0x%x\n", + __func__, props->cpu_family); + client->revision = 0x31; + break; + } + + qdev_realize_and_unref(&client->qdev, &bus->qbus, &error_abort); + bus->num_clients += 1; + return client; +} + +static void peci_client_get(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + /* use millidegrees convention */ + uint32_t value = *(uint8_t *)opaque * 1000; + visit_type_uint32(v, name, &value, errp); +} + +static void peci_client_set(Object *obj, Visitor *v, const char *name, + void *opaque, Error **errp) +{ + PECIClientDevice *client = PECI_CLIENT(obj); + uint8_t *internal = opaque; + uint32_t value; + + if (!visit_type_uint32(v, name, &value, errp)) { + return; + } + + g_assert(value <= 255000); + + *internal = value / 1000; + peci_client_update_temps(client); +} + +static void peci_client_reset(Object *obj, ResetType type) +{ + PECIClientDevice *client = PECI_CLIENT(obj); + client->core_temp_max = 0; + client->dimm_temp_max = 0; +} + +static const VMStateDescription vmstate_peci_client = { + .name = TYPE_PECI_CLIENT, + .version_id = 0, + .minimum_version_id = 0, + .fields = (VMStateField[]) { + VMSTATE_UINT8(address, PECIClientDevice), + VMSTATE_UINT8(device_info, PECIClientDevice), + VMSTATE_UINT8(revision, PECIClientDevice), + VMSTATE_UINT32(cpu_id, PECIClientDevice), + VMSTATE_UINT8(cpu_cores, PECIClientDevice), + VMSTATE_UINT8(tjmax, PECIClientDevice), + VMSTATE_INT8(tcontrol, PECIClientDevice), + VMSTATE_UINT8_ARRAY(core_temp, PECIClientDevice, PECI_NUM_CPUS_MAX), + VMSTATE_UINT8(dimms, PECIClientDevice), + VMSTATE_UINT8(dimms_per_channel, PECIClientDevice), + VMSTATE_UINT8_ARRAY(dimm_temp, PECIClientDevice , PECI_NUM_DIMMS_MAX), + VMSTATE_END_OF_LIST() + }, +}; + +static void peci_client_realize(DeviceState *dev, Error **errp) +{ + PECIClientDevice *client = PECI_CLIENT(dev); + for (size_t i = 0; i < client->cpu_cores; i++) { + client->core_temp[i] = PECI_CLIENT_DEFAULT_TEMP; + object_property_add(OBJECT(client), "cpu_temp[*]", "uint8", + peci_client_get, + peci_client_set, NULL, &client->core_temp[i]); + } + + for (size_t i = 0; i < client->dimms; i++) { + client->dimm_temp[i] = PECI_CLIENT_DEFAULT_TEMP; + object_property_add(OBJECT(client), "dimm_temp[*]", "uint8", + peci_client_get, + peci_client_set, NULL, &client->dimm_temp[i]); + } +} + +static void peci_client_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + ResettableClass *rc = RESETTABLE_CLASS(klass); + + dc->desc = "PECI Client Device"; + dc->bus_type = TYPE_PECI_BUS; + dc->realize = peci_client_realize; + dc->vmsd = &vmstate_peci_client; + set_bit(DEVICE_CATEGORY_MISC, dc->categories); + rc->phases.enter = peci_client_reset; +} + +static const TypeInfo peci_client_info = { + .name = TYPE_PECI_CLIENT, + .parent = TYPE_DEVICE, + .instance_size = sizeof(PECIClientDevice), + .class_init = peci_client_class_init, +}; + +static void peci_client_register_types(void) +{ + type_register_static(&peci_client_info); +} + +type_init(peci_client_register_types) diff --git a/hw/peci/peci-core.c b/hw/peci/peci-core.c new file mode 100644 index 0000000000..8210bfa198 --- /dev/null +++ b/hw/peci/peci-core.c @@ -0,0 +1,182 @@ +/* + * PECI - Platform Environment Control Interface + * Copyright 2021 Google LLC + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +/* + * This PECI implementation aims to simulate a host with peci as viewed by a + * BMC. This is developed with OpenBMC firmware running in QEMU. + * + */ + +#include "qemu/osdep.h" +#include "hw/peci/peci.h" +#include "hw/qdev-properties.h" +#include "migration/vmstate.h" +#include "qemu/module.h" +#include "qemu/log.h" +#include "trace.h" + +#define PECI_FCS_OK 0 +#define PECI_FCS_ERR 1 + +static void peci_rd_pkg_cfg(PECIClientDevice *client, PECICmd *pcmd) +{ + PECIPkgCfg *resp = (PECIPkgCfg *)pcmd->tx; + uint8_t index = pcmd->rx[1]; + uint16_t param = pcmd->rx[3] | pcmd->rx[2]; + + switch (index) { + case PECI_MBX_CPU_ID: /* CPU Family ID*/ + trace_peci_rd_pkg_cfg("CPU ID"); + switch (param) { + case PECI_PKG_ID_CPU_ID: + resp->data = client->cpu_id; + break; + + case PECI_PKG_ID_MICROCODE_REV: + resp->data = client->ucode; + break; + + default: + qemu_log_mask(LOG_UNIMP, "%s: CPU ID param %u\n", __func__, param); + break; + } + break; + + case PECI_MBX_DTS_MARGIN: + trace_peci_rd_pkg_cfg("DTS MARGIN"); + /* + * Processors return a value of DTS reading in 10.6 format + * (10 bits signed decimal, 6 bits fractional). + */ + resp->data = (-1 * client->core_temp_max) << 6; + break; + + case PECI_MBX_DDR_DIMM_TEMP: + trace_peci_rd_pkg_cfg("DDR DIMM TEMP"); + if (param > PECI_NUM_CHANNELS_MAX) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: DDR_DIMM_TEMP unsupported channel index %d", + __func__, param); + param = PECI_NUM_CHANNELS_MAX; + } + uint8_t channel_index = param * client->dimms_per_channel; + memcpy(&resp->data, + &client->dimm_temp[channel_index], + sizeof(client->dimm_temp[0]) * client->dimms_per_channel); + break; + + case PECI_MBX_TEMP_TARGET: + trace_peci_rd_pkg_cfg("TEMP TARGET"); + PECITempTarget target = { + .tcontrol = client->tcontrol, + .tjmax = client->tjmax, + }; + memcpy(&resp->data, &target, sizeof(target)); + break; + + case PECI_MBX_PPIN: + trace_peci_rd_pkg_cfg("PPIN"); + resp->data = 0xdeadbeef; + break; + + default: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented PkgCfg Index: %u\n", + __func__, index); + resp->cc = PECI_DEV_CC_INVALID_REQ; + return; + } + + resp->cc = PECI_DEV_CC_SUCCESS; +} + +int peci_handle_cmd(PECIBus *bus, PECICmd *pcmd) +{ + PECIClientDevice *client = peci_get_client(bus, pcmd->addr); + + if (!client) { /* ignore commands if clients aren't found */ + qemu_log_mask(LOG_GUEST_ERROR, + "%s: no peci client at address: 0x%02x\n", + __func__, pcmd->addr); + return PECI_FCS_ERR; + } + + /* clear output buffer before handling command */ + for (size_t i = 0; i < pcmd->rd_length; i++) { + pcmd->tx[i] = 0; + } + + switch (pcmd->cmd) { + case PECI_CMD_PING: + trace_peci_handle_cmd("PING!", pcmd->addr); + break; + + case PECI_CMD_GET_DIB: /* Device Info Byte */ + trace_peci_handle_cmd("GetDIB", pcmd->addr); + PECIDIB dib = { + .device_info = client->device_info, + .revision = client->revision, + }; + memcpy(pcmd->tx, &dib, sizeof(PECIDIB)); + break; + + case PECI_CMD_GET_TEMP: /* maximum die temp in socket */ + trace_peci_handle_cmd("GetTemp", pcmd->addr); + /* + * The data is returned as a negative value representing the number of + * degrees centigrade below the maximum processor junction temperature + */ + memcpy(pcmd->tx, &client->core_temp_max, sizeof(client->core_temp_max)); + break; + + case PECI_CMD_RD_PKG_CFG: + trace_peci_handle_cmd("RdPkgConfig", pcmd->addr); + peci_rd_pkg_cfg(client, pcmd); + break; + + case PECI_CMD_RD_IA_MSR: + trace_peci_handle_cmd("RdIAMSR", pcmd->addr); + qemu_log_mask(LOG_UNIMP, "%s: unimplemented RD_IA_MSR\n", __func__); + break; + + case PECI_CMD_RD_PCI_CFG: + trace_peci_handle_cmd("RdPCIConfig", pcmd->addr); + qemu_log_mask(LOG_UNIMP, "%s: unimplemented PCI_CFG\n", __func__); + break; + + case PECI_CMD_RD_PCI_CFG_LOCAL: + trace_peci_handle_cmd("RdPCIConfigLocal", pcmd->addr); + qemu_log_mask(LOG_UNIMP, "%s: unimplemented PCI_CFG_LOCAL\n", __func__); + break; + + case PECI_CMD_RD_END_PT_CFG: + qemu_log_mask(LOG_UNIMP, "%s: unimplemented CMD_RD_END_PT_CFG\n", + __func__); + break; + + default: + qemu_log_mask(LOG_GUEST_ERROR, "%s: unknown cmd: %x\n", + __func__, pcmd->cmd); + break; + } + + return PECI_FCS_OK; +} + +PECIBus *peci_bus_create(DeviceState *parent) +{ + return PECI_BUS(qbus_new(TYPE_PECI_BUS, parent, "peci-bus")); +}; + +static const TypeInfo peci_types[] = { + { + .name = TYPE_PECI_BUS, + .parent = TYPE_BUS, + .instance_size = sizeof(PECIBus), + }, +}; + +DEFINE_TYPES(peci_types); diff --git a/hw/peci/trace-events b/hw/peci/trace-events new file mode 100644 index 0000000000..f90c998dd9 --- /dev/null +++ b/hw/peci/trace-events @@ -0,0 +1,5 @@ +# See docs/devel/tracing.rst for syntax documentation. + +# peci-core.c +peci_handle_cmd(const char* s, uint8_t addr) "%s @ 0x%02" PRIx8 +peci_rd_pkg_cfg(const char* s) "%s" diff --git a/hw/peci/trace.h b/hw/peci/trace.h new file mode 100644 index 0000000000..4283e3bfc5 --- /dev/null +++ b/hw/peci/trace.h @@ -0,0 +1 @@ +#include "trace/trace-hw_peci.h" diff --git a/include/hw/peci/peci.h b/include/hw/peci/peci.h new file mode 100644 index 0000000000..1a0abe65cd --- /dev/null +++ b/include/hw/peci/peci.h @@ -0,0 +1,194 @@ +/* + * PECI - Platform Environment Control Interface + * + * Copyright 2021 Google LLC + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#ifndef PECI_H +#define PECI_H + +#include "hw/qdev-core.h" +#include "qom/object.h" + +#define PECI_CMD_PING 0x0 +#define PECI_CMD_GET_DIB 0xF7 +#define PECI_CMD_GET_TEMP 0x01 +#define PECI_CMD_RD_PKG_CFG 0xA1 +#define PECI_CMD_WR_PKG_CFG 0xA5 +#define PECI_MBX_CPU_ID 0 /* Package Identifier Read */ +#define PECI_PKG_ID_CPU_ID 0x0000 /* CPUID Info */ +#define PECI_PKG_ID_PLATFORM_ID 0x0001 /* Platform ID */ +#define PECI_PKG_ID_UNCORE_ID 0x0002 /* Uncore Device ID */ +#define PECI_PKG_ID_MAX_THREAD_ID 0x0003 /* Max Thread ID */ +#define PECI_PKG_ID_MICROCODE_REV 0x0004 /* CPU Microcode Revision */ +#define PECI_PKG_ID_MACHINE_CHECK_STATUS 0x0005 /* Machine Check Status */ +#define PECI_MBX_VR_DEBUG 1 /* VR Debug */ +#define PECI_MBX_PKG_TEMP_READ 2 /* Package Temperature Read */ +#define PECI_MBX_ENERGY_COUNTER 3 /* Energy counter */ +#define PECI_MBX_ENERGY_STATUS 4 /* DDR Energy Status */ +#define PECI_MBX_WAKE_MODE_BIT 5 /* "Wake on PECI" Mode bit */ +#define PECI_MBX_EPI 6 /* Efficient Performance Indication */ +#define PECI_MBX_PKG_RAPL_PERF 8 /* Pkg RAPL Performance Status Read */ +#define PECI_MBX_PER_CORE_DTS_TEMP 9 /* Per Core DTS Temperature Read */ +#define PECI_MBX_DTS_MARGIN 10 /* DTS thermal margin */ +#define PECI_MBX_SKT_PWR_THRTL_DUR 11 /* Socket Power Throttled Duration */ +#define PECI_MBX_CFG_TDP_CONTROL 12 /* TDP Config Control */ +#define PECI_MBX_CFG_TDP_LEVELS 13 /* TDP Config Levels */ +#define PECI_MBX_DDR_DIMM_TEMP 14 /* DDR DIMM Temperature */ +#define PECI_MBX_CFG_ICCMAX 15 /* Configurable ICCMAX */ +#define PECI_MBX_TEMP_TARGET 16 /* Temperature Target Read */ +#define PECI_MBX_CURR_CFG_LIMIT 17 /* Current Config Limit */ +#define PECI_MBX_PPIN 19 /* Processor Inventory Number */ +#define PECI_MBX_DIMM_TEMP_READ 20 /* Package Thermal Status Read */ +#define PECI_MBX_DRAM_IMC_TMP_READ 22 /* DRAM IMC Temperature Read */ +#define PECI_MBX_DDR_CH_THERM_STAT 23 /* DDR Channel Thermal Status */ +#define PECI_MBX_PKG_POWER_LIMIT1 26 /* Package Power Limit1 */ +#define PECI_MBX_PKG_POWER_LIMIT2 27 /* Package Power Limit2 */ +#define PECI_MBX_TDP 28 /* Thermal design power minimum */ +#define PECI_MBX_TDP_HIGH 29 /* Thermal design power maximum */ +#define PECI_MBX_TDP_UNITS 30 /* Units for power/energy registers */ +#define PECI_MBX_RUN_TIME 31 /* Accumulated Run Time */ +#define PECI_MBX_CONSTRAINED_TIME 32 /* Thermally Constrained Time Read */ +#define PECI_MBX_TURBO_RATIO 33 /* Turbo Activation Ratio */ +#define PECI_MBX_DDR_RAPL_PL1 34 /* DDR RAPL PL1 */ +#define PECI_MBX_DDR_PWR_INFO_HIGH 35 /* DRAM Power Info Read (high) */ +#define PECI_MBX_DDR_PWR_INFO_LOW 36 /* DRAM Power Info Read (low) */ +#define PECI_MBX_DDR_RAPL_PL2 37 /* DDR RAPL PL2 */ +#define PECI_MBX_DDR_RAPL_STATUS 38 /* DDR RAPL Performance Status */ +#define PECI_MBX_DDR_HOT_ABSOLUTE 43 /* DDR Hottest Dimm Absolute Temp */ +#define PECI_MBX_DDR_HOT_RELATIVE 44 /* DDR Hottest Dimm Relative Temp */ +#define PECI_MBX_DDR_THROTTLE_TIME 45 /* DDR Throttle Time */ +#define PECI_MBX_DDR_THERM_STATUS 46 /* DDR Thermal Status */ +#define PECI_MBX_TIME_AVG_TEMP 47 /* Package time-averaged temperature */ +#define PECI_MBX_TURBO_RATIO_LIMIT 49 /* Turbo Ratio Limit Read */ +#define PECI_MBX_HWP_AUTO_OOB 53 /* HWP Autonomous Out-of-band */ +#define PECI_MBX_DDR_WARM_BUDGET 55 /* DDR Warm Power Budget */ +#define PECI_MBX_DDR_HOT_BUDGET 56 /* DDR Hot Power Budget */ +#define PECI_MBX_PKG_PSYS_PWR_LIM3 57 /* Package/Psys Power Limit3 */ +#define PECI_MBX_PKG_PSYS_PWR_LIM1 58 /* Package/Psys Power Limit1 */ +#define PECI_MBX_PKG_PSYS_PWR_LIM2 59 /* Package/Psys Power Limit2 */ +#define PECI_MBX_PKG_PSYS_PWR_LIM4 60 /* Package/Psys Power Limit4 */ +#define PECI_MBX_PERF_LIMIT_REASON 65 /* Performance Limit Reasons */ +#define PECI_CMD_RD_IA_MSR 0xB1 +#define PECI_CMD_WR_IA_MSR 0xB5 +#define PECI_CMD_RD_IA_MSREX 0xD1 +#define PECI_CMD_RD_PCI_CFG 0x61 +#define PECI_CMD_WR_PCI_CFG 0x65 +#define PECI_CMD_RD_PCI_CFG_LOCAL 0xE1 +#define PECI_CMD_WR_PCI_CFG_LOCAL 0xE5 +#define PECI_CMD_RD_END_PT_CFG 0xC1 +#define PECI_CMD_WR_END_PT_CFG 0xC5 +#define PECI_CMD_CRASHDUMP_GET_FRAME 0x71 + +/* Device Specific Completion Code (CC) Definition */ +#define PECI_DEV_CC_SUCCESS 0x40 +#define PECI_DEV_CC_NEED_RETRY 0x80 +#define PECI_DEV_CC_OUT_OF_RESOURCE 0x81 +#define PECI_DEV_CC_UNAVAIL_RESOURCE 0x82 +#define PECI_DEV_CC_INVALID_REQ 0x90 +#define PECI_DEV_CC_MCA_ERROR 0x91 +#define PECI_DEV_CC_CATASTROPHIC_MCA_ERROR 0x93 +#define PECI_DEV_CC_FATAL_MCA_DETECTED 0x94 +#define PECI_DEV_CC_PARITY_ERROR_ON_GPSB_OR_PMSB 0x98 +#define PECI_DEV_CC_PARITY_ERROR_ON_GPSB_OR_PMSB_IERR 0x9B +#define PECI_DEV_CC_PARITY_ERROR_ON_GPSB_OR_PMSB_MCA 0x9C + + +typedef struct PECIDIB { /* DIB - Device Info Byte(s) */ + uint8_t device_info; /* bit 2 - num of domains */ + uint8_t revision; +} PECIDIB; + +typedef struct __attribute__ ((__packed__)) { + uint8_t cc; /* completion code */ + uint32_t data; +} PECIPkgCfg; + +typedef struct PECITempTarget { + uint8_t reserved; + int8_t tcontrol; + uint8_t tjmax; +} PECITempTarget; + +#define PECI_BASE_ADDR 0x30 +#define PECI_BUFFER_SIZE 0x100 +#define PECI_NUM_CPUS_MAX 56 +#define PECI_DIMMS_PER_CHANNEL_MAX 3 +#define PECI_NUM_CHANNELS_MAX 8 +#define PECI_NUM_DIMMS_MAX (PECI_NUM_CHANNELS_MAX * PECI_DIMMS_PER_CHANNEL_MAX) + +typedef struct PECIClientDevice { + DeviceState qdev; + uint8_t address; + uint8_t device_info; + uint8_t revision; + + /* CPU */ + uint32_t cpu_id; + uint8_t cpu_cores; + uint32_t ucode; + uint8_t tjmax; + int8_t tcontrol; + int8_t core_temp_max; /* absolute temp - tjmax */ + uint8_t core_temp[PECI_NUM_CPUS_MAX]; + + /* DIMMS */ + uint8_t dimms; + uint8_t dimms_per_channel; + uint8_t dimm_temp_max; + uint8_t dimm_temp[PECI_NUM_DIMMS_MAX]; + +} PECIClientDevice; + +#define TYPE_PECI_CLIENT "peci-client" +OBJECT_DECLARE_SIMPLE_TYPE(PECIClientDevice, PECI_CLIENT) + +typedef struct PECIBus { + BusState qbus; + uint8_t num_clients; +} PECIBus; + +#define TYPE_PECI_BUS "peci-bus" +OBJECT_DECLARE_SIMPLE_TYPE(PECIBus, PECI_BUS) + +/* Creates a basic qemu bus onto which PECI clients will be attached */ +PECIBus *peci_bus_create(DeviceState *parent); + +enum { + FAM6_HASWELL_X = 0x306F0, + FAM6_BROADWELL_X = 0x406F0, + FAM6_SKYLAKE_X = 0x50650, + FAM6_SKYLAKE_XD = 0x50660, + FAM6_ICELAKE_X = 0x606A0, + FAM6_SAPPHIRE_RAPIDS_X = 0x806F3, +}; + +typedef struct PECIClientProperties { + uint32_t cpu_family; + uint8_t cpu_cores; + uint8_t dimms; + uint8_t dimms_per_channel; +} PECIClientProperties; + +/* Creates a PECI client with the specified cpu and dimm count */ +PECIClientDevice *peci_add_client(PECIBus *bus, + uint8_t address, + PECIClientProperties *props); +PECIClientDevice *peci_get_client(PECIBus *bus, uint8_t addr); + +typedef struct PECICmd { + uint8_t addr; + uint8_t cmd; + + size_t wr_length; + uint8_t rx[PECI_BUFFER_SIZE]; + + size_t rd_length; + uint8_t tx[PECI_BUFFER_SIZE]; +} PECICmd; + +int peci_handle_cmd(PECIBus *bus, PECICmd *pcmd); + +#endif diff --git a/meson.build b/meson.build index 20fddbd707..55594702c3 100644 --- a/meson.build +++ b/meson.build @@ -2965,6 +2965,7 @@ if have_system 'hw/nvram', 'hw/pci', 'hw/pci-host', + 'hw/peci', 'hw/ppc', 'hw/rdma', 'hw/rdma/vmw',