[RFC PATCH v2 0/2] PCIe DOE for PCIe and CXL 2.0 v2 release

[RFC PATCH v2 0/2] PCIe DOE for PCIe and CXL 2.0 v2 release

[Chris Browy]

Version 2 patch series for PCIe DOE for PCIe and CXL 2.0

Summary is 

1: PCIe DOE support for Discovery and CMA.
   - MSI-X and polling supported
2: CXL DOE for CDAT and Compliance Mode.
   - DOE CDAT response returns one CDAT Structure instance based on
     request EntryHandle value.
   - One of each CDAT Structure types supported

Based on QEMU version:
https://gitlab.com/bwidawsk/qemu/-/tree/cxl-2.0v3

References:
1. CXL 2.0 specification
https://www.computeexpresslink.org/download-the-specification
2. PCI-SIG ECN: Data Object Exchange (DOE)
http://www.pcisig.com
3. Coherent Device Attribute Table	CDAT 1.02
https://uefi.org/sites/default/files/resources/Coherent%20Device%20Attribute%20Table_1.02.pdf

---

Chris Browy (2):
  Basic PCIe DOE support
  Basic CXL DOE for CDAT and Compliance Mode

 MAINTAINERS                               |   7 +
 hw/cxl/cxl-component-utils.c              | 132 ++++++++++
 hw/mem/cxl_type3.c                        | 172 +++++++++++++
 hw/pci/meson.build                        |   1 +
 hw/pci/pcie.c                             |   2 +-
 hw/pci/pcie_doe.c                         | 414 ++++++++++++++++++++++++++++++
 include/hw/cxl/cxl_cdat.h                 | 120 +++++++++
 include/hw/cxl/cxl_compl.h                | 289 +++++++++++++++++++++
 include/hw/cxl/cxl_component.h            | 126 +++++++++
 include/hw/cxl/cxl_device.h               |   3 +
 include/hw/cxl/cxl_pci.h                  |   4 +
 include/hw/pci/pci_ids.h                  |   2 +
 include/hw/pci/pcie.h                     |   1 +
 include/hw/pci/pcie_doe.h                 | 166 ++++++++++++
 include/hw/pci/pcie_regs.h                |   4 +
 include/standard-headers/linux/pci_regs.h |   3 +-
 16 files changed, 1444 insertions(+), 2 deletions(-)
 create mode 100644 hw/pci/pcie_doe.c
 create mode 100644 include/hw/cxl/cxl_cdat.h
 create mode 100644 include/hw/cxl/cxl_compl.h
 create mode 100644 include/hw/pci/pcie_doe.h

-- 
1.8.3.1

[RFC PATCH v2 1/2] Basic PCIe DOE support

[Chris Browy]

---
 MAINTAINERS                               |   7 +
 hw/pci/meson.build                        |   1 +
 hw/pci/pcie.c                             |   2 +-
 hw/pci/pcie_doe.c                         | 414 ++++++++++++++++++++++++++++++
 include/hw/pci/pci_ids.h                  |   2 +
 include/hw/pci/pcie.h                     |   1 +
 include/hw/pci/pcie_doe.h                 | 166 ++++++++++++
 include/hw/pci/pcie_regs.h                |   4 +
 include/standard-headers/linux/pci_regs.h |   3 +-
 9 files changed, 598 insertions(+), 2 deletions(-)
 create mode 100644 hw/pci/pcie_doe.c
 create mode 100644 include/hw/pci/pcie_doe.h

diff --git a/MAINTAINERS b/MAINTAINERS
index 981dc92..4fb865e 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1655,6 +1655,13 @@ F: docs/pci*
 F: docs/specs/*pci*
 F: default-configs/pci.mak
 
+PCIE DOE
+M: Huai-Cheng Kuo <hchkuo@avery-design.com.tw>
+M: Chris Browy <cbrowy@avery-design.com>
+S: Supported
+F: include/hw/pci/pcie_doe.h
+F: hw/pci/pcie_doe.c
+
 ACPI/SMBIOS
 M: Michael S. Tsirkin <mst@redhat.com>
 M: Igor Mammedov <imammedo@redhat.com>
diff --git a/hw/pci/meson.build b/hw/pci/meson.build
index 5c4bbac..115e502 100644
--- a/hw/pci/meson.build
+++ b/hw/pci/meson.build
@@ -12,6 +12,7 @@ pci_ss.add(files(
 # allow plugging PCIe devices into PCI buses, include them even if
 # CONFIG_PCI_EXPRESS=n.
 pci_ss.add(files('pcie.c', 'pcie_aer.c'))
+pci_ss.add(files('pcie_doe.c'))
 softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: files('pcie_port.c', 'pcie_host.c'))
 softmmu_ss.add_all(when: 'CONFIG_PCI', if_true: pci_ss)
 
diff --git a/hw/pci/pcie.c b/hw/pci/pcie.c
index 1ecf6f6..f7516c4 100644
--- a/hw/pci/pcie.c
+++ b/hw/pci/pcie.c
@@ -735,7 +735,7 @@ void pcie_cap_slot_write_config(PCIDevice *dev,
 
     hotplug_event_notify(dev);
 
-    /* 
+    /*
      * 6.7.3.2 Command Completed Events
      *
      * Software issues a command to a hot-plug capable Downstream Port by
diff --git a/hw/pci/pcie_doe.c b/hw/pci/pcie_doe.c
new file mode 100644
index 0000000..df8e92e
--- /dev/null
+++ b/hw/pci/pcie_doe.c
@@ -0,0 +1,414 @@
+#include "qemu/osdep.h"
+#include "qemu/log.h"
+#include "qemu/error-report.h"
+#include "qapi/error.h"
+#include "qemu/range.h"
+#include "hw/pci/pci.h"
+#include "hw/pci/pcie.h"
+#include "hw/pci/pcie_doe.h"
+#include "hw/pci/msi.h"
+#include "hw/pci/msix.h"
+
+/*
+ * DOE Default Protocols (Discovery, CMA)
+ */
+/* Discovery Request Object */
+struct doe_discovery {
+    DOEHeader header;
+    uint8_t index;
+    uint8_t reserved[3];
+} QEMU_PACKED;
+
+/* Discovery Response Object */
+struct doe_discovery_rsp {
+    DOEHeader header;
+    uint16_t vendor_id;
+    uint8_t doe_type;
+    uint8_t next_index;
+} QEMU_PACKED;
+
+/* Callback for Discovery */
+static bool pcie_doe_discovery_rsp(DOECap *doe_cap)
+{
+    PCIEDOE *doe = doe_cap->doe;
+    struct doe_discovery *req = pcie_doe_get_req(doe_cap);
+    uint8_t index = req->index;
+    DOEProtocol *prot = NULL;
+
+    /* Request length mismatch, discard */
+    if (req->header.length < dwsizeof(struct doe_discovery)) {
+        return DOE_DISCARD;
+    }
+
+    /* Point to the requested protocol */
+    if (index < doe->protocol_num) {
+        prot = &doe->protocols[index];
+    }
+
+    struct doe_discovery_rsp rsp = {
+        .header = {
+            .vendor_id = PCI_VENDOR_ID_PCI_SIG,
+            .doe_type = PCI_SIG_DOE_DISCOVERY,
+            .reserved = 0x0,
+            .length = dwsizeof(struct doe_discovery_rsp),
+        },
+        .vendor_id = (prot) ? prot->vendor_id : 0xFFFF,
+        .doe_type = (prot) ? prot->doe_type : 0xFF,
+        .next_index = (index + 1) < doe->protocol_num ?
+                      (index + 1) : 0,
+    };
+
+    pcie_doe_set_rsp(doe_cap, &rsp);
+
+    return DOE_SUCCESS;
+}
+
+/* Callback for CMA */
+static bool pcie_doe_cma_rsp(DOECap *doe_cap)
+{
+    doe_cap->status.error = 1;
+
+    memset(doe_cap->read_mbox, 0,
+           PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
+
+    doe_cap->write_mbox_len = 0;
+
+    return DOE_DISCARD;
+}
+
+/*
+ * DOE Utilities
+ */
+static void pcie_doe_reset_mbox(DOECap *st)
+{
+    st->read_mbox_idx = 0;
+
+    st->read_mbox_len = 0;
+    st->write_mbox_len = 0;
+
+    memset(st->read_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
+    memset(st->write_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
+}
+
+/*
+ * Initialize the list and protocol for a device.
+ * This function won't add the DOE capabitity to your PCIe device.
+ */
+void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe)
+{
+    doe->pdev = dev;
+    doe->head = NULL;
+    doe->protocol_num = 0;
+
+    /* Register two default protocol */
+    //TODO : LINK LIST
+    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
+            PCI_SIG_DOE_DISCOVERY, pcie_doe_discovery_rsp);
+    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
+            PCI_SIG_DOE_CMA, pcie_doe_cma_rsp);
+}
+
+int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec) {
+    DOECap *new_cap, **ptr;
+    PCIDevice *dev = doe->pdev;
+
+    pcie_add_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_VER, offset,
+                        PCI_DOE_SIZEOF);
+
+    ptr = &doe->head;
+    /* Insert the new DOE at the end of linked list */
+    while (*ptr) {
+        if (range_covers_byte((*ptr)->offset, PCI_DOE_SIZEOF, offset) ||
+            (*ptr)->cap.vec == vec) {
+            return -EINVAL;
+        }
+
+        ptr = &(*ptr)->next;
+    }
+    new_cap = g_malloc0(sizeof(DOECap));
+    *ptr = new_cap;
+
+    new_cap->doe = doe;
+    new_cap->offset = offset;
+    new_cap->next = NULL;
+
+    /* Configure MSI/MSI-X */
+    if (intr && (msi_present(dev) || msix_present(dev))) {
+        new_cap->cap.intr = intr;
+        new_cap->cap.vec = vec;
+    }
+
+    /* Set up W/R Mailbox buffer */
+    new_cap->write_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
+    new_cap->read_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
+
+    pcie_doe_reset_mbox(new_cap);
+
+    return 0;
+}
+
+void pcie_doe_uninit(PCIEDOE *doe) {
+    PCIDevice *dev = doe->pdev;
+    DOECap *cap;
+
+    pci_del_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_SIZEOF);
+
+    cap = doe->head;
+    while (cap) {
+        doe->head = doe->head->next;
+
+        g_free(cap->read_mbox);
+        g_free(cap->write_mbox);
+        cap->read_mbox = NULL;
+        cap->write_mbox = NULL;
+        g_free(cap);
+        cap = doe->head;
+    }
+}
+
+void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
+        uint8_t doe_type, bool (*set_rsp)(DOECap *))
+{
+    /* Protocol num should not exceed 256 */
+    assert(doe->protocol_num < PCI_DOE_PROTOCOL_MAX);
+
+    doe->protocols[doe->protocol_num].vendor_id = vendor_id;
+    doe->protocols[doe->protocol_num].doe_type = doe_type;
+    doe->protocols[doe->protocol_num].set_rsp = set_rsp;
+
+    doe->protocol_num++;
+}
+
+uint32_t pcie_doe_build_protocol(DOEProtocol *p)
+{
+    return DATA_OBJ_BUILD_HEADER1(p->vendor_id, p->doe_type);
+}
+
+/* Return the pointer of DOE request in write mailbox buffer */
+void *pcie_doe_get_req(DOECap *doe_cap)
+{
+    return doe_cap->write_mbox;
+}
+
+/* Copy the response to read mailbox buffer */
+void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp)
+{
+    uint32_t len = pcie_doe_object_len(rsp);
+
+    memcpy(doe_cap->read_mbox + doe_cap->read_mbox_len,
+           rsp, len * sizeof(uint32_t));
+    doe_cap->read_mbox_len += len;
+}
+
+/* Get Data Object length */
+uint32_t pcie_doe_object_len(void *obj)
+{
+    return (obj)? ((DOEHeader *)obj)->length : 0;
+}
+
+static void pcie_doe_write_mbox(DOECap *doe_cap, uint32_t val)
+{
+    memcpy(doe_cap->write_mbox + doe_cap->write_mbox_len, &val, sizeof(uint32_t));
+
+    if (doe_cap->write_mbox_len == 1) {
+        DOE_DBG("  Capture DOE request DO length = %d\n", val & 0x0003ffff);
+    }
+
+    doe_cap->write_mbox_len++;
+}
+
+static void pcie_doe_irq_assert(DOECap *doe_cap)
+{
+    PCIDevice *dev = doe_cap->doe->pdev;
+
+    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {
+        /* Interrupt notify */
+        if (msix_enabled(dev)) {
+            msix_notify(dev, doe_cap->cap.vec);
+        } else if (msi_enabled(dev)) {
+            msi_notify(dev, doe_cap->cap.vec);
+        }
+        /* Not support legacy IRQ */
+    }
+}
+
+static void pcie_doe_set_ready(DOECap *doe_cap, bool rdy)
+{
+    doe_cap->status.ready = rdy;
+
+    if (rdy) {
+        pcie_doe_irq_assert(doe_cap);
+    }
+}
+
+static void pcie_doe_set_error(DOECap *doe_cap, bool err)
+{
+    doe_cap->status.error = err;
+
+    if (err) {
+        pcie_doe_irq_assert(doe_cap);
+    }
+}
+
+DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr)
+{
+    DOECap *ptr = doe->head;
+
+    /* If overlaps DOE range. PCIe Capability Header won't be included. */
+    while (ptr && 
+           !range_covers_byte(ptr->offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
+        ptr = ptr->next;
+    }
+    
+    return ptr;
+}
+
+uint32_t pcie_doe_read_config(DOECap *doe_cap,
+                              uint32_t addr, int size)
+{
+    uint32_t ret = 0, shift, mask = 0xFFFFFFFF;
+    uint16_t doe_offset = doe_cap->offset;
+
+    /* If overlaps DOE range. PCIe Capability Header won't be included. */
+    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
+        addr -= doe_offset;
+
+        if (range_covers_byte(PCI_EXP_DOE_CAP, sizeof(uint32_t), addr)) {
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, INTR_SUPP,
+                             doe_cap->cap.intr);
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM,
+                             doe_cap->cap.vec);
+        } else if (range_covers_byte(PCI_EXP_DOE_CTRL, sizeof(uint32_t), addr)) {
+            /* Must return ABORT=0 and GO=0 */
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_CONTROL, DOE_INTR_EN,
+                             doe_cap->ctrl.intr);
+            DOE_DBG("  CONTROL REG=%x\n", ret);
+        } else if (range_covers_byte(PCI_EXP_DOE_STATUS, sizeof(uint32_t), addr)) {
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_BUSY,
+                             doe_cap->status.busy);
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_INTR_STATUS,
+                             doe_cap->status.intr);
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_ERROR,
+                             doe_cap->status.error);
+            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DATA_OBJ_RDY,
+                             doe_cap->status.ready);
+        } else if (range_covers_byte(PCI_EXP_DOE_RD_DATA_MBOX, sizeof(uint32_t), addr)) {
+            /* Check that READY is set */
+            if (!doe_cap->status.ready) {
+                /* Return 0 if not ready */
+                DOE_DBG("  RD MBOX RETURN=%x\n", ret);
+            } else {
+                /* Deposit next DO DW into read mbox */
+                DOE_DBG("  RD MBOX, DATA OBJECT READY,"
+                        " CURRENT DO DW OFFSET=%x\n",
+                        doe_cap->read_mbox_idx);
+
+                ret = doe_cap->read_mbox[doe_cap->read_mbox_idx];
+
+                DOE_DBG("  RD MBOX DW=%x\n", ret);
+                DOE_DBG("  RD MBOX DW OFFSET=%d, RD MBOX LENGTH=%d\n",
+                        doe_cap->read_mbox_idx, doe_cap->read_mbox_len);
+            }
+        } else if (range_covers_byte(PCI_EXP_DOE_WR_DATA_MBOX, sizeof(uint32_t), addr)) {
+            DOE_DBG("  WR MBOX, local_val =%x\n", ret);
+        }
+    }
+
+    /* Alignment */
+    shift = addr % sizeof(uint32_t);
+    if (shift) {
+        ret >>= shift * 8;
+    }
+    mask >>= (sizeof(uint32_t) - size) * 8;
+
+    return ret & mask;
+}
+
+void pcie_doe_write_config(DOECap *doe_cap,
+                           uint32_t addr, uint32_t val, int size)
+{
+    PCIEDOE *doe = doe_cap->doe;
+    uint16_t doe_offset = doe_cap->offset;
+    int p;
+    bool discard;
+    uint32_t shift;
+
+    DOE_DBG("  addr=%x, val=%x, size=%x\n", addr, val, size);
+
+    /* If overlaps DOE range. PCIe Capability Header won't be included. */
+    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
+        /* Alignment */
+        shift = addr % sizeof(uint32_t);
+        addr -= (doe_offset - shift);
+        val <<= shift * 8;
+
+        switch (addr) {
+        case PCI_EXP_DOE_CTRL:
+            DOE_DBG("  CONTROL=%x\n", val);
+            if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_ABORT)) {
+                /* If ABORT, clear status reg DOE Error and DOE Ready */
+                DOE_DBG("  Setting ABORT\n");
+                pcie_doe_set_ready(doe_cap, 0);
+                pcie_doe_set_error(doe_cap, 0);
+                pcie_doe_reset_mbox(doe_cap);
+            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_GO)) {
+                DOE_DBG("  CONTROL GO=%x\n", val);
+                /*
+                 * Check protocol the incoming request in write_mbox and
+                 * memcpy the corresponding response to read_mbox.
+                 *
+                 * "discard" should be set up if the response callback
+                 * function could not deal with request (e.g. length
+                 * mismatch) or the protocol of request was not found.
+                 */
+                discard = DOE_DISCARD;
+                for (p = 0; p < doe->protocol_num; p++) {
+                    if (doe_cap->write_mbox[0] ==
+                        pcie_doe_build_protocol(&doe->protocols[p])) {
+                        /* Found */
+                        DOE_DBG("  DOE PROTOCOL=%x\n", doe_cap->write_mbox[0]);
+                        /*
+                         * Spec:
+                         * If the number of DW transferred does not match the
+                         * indicated Length for a data object, then the
+                         * data object must be silently discarded.
+                         */
+                        if (doe_cap->write_mbox_len ==
+                            pcie_doe_object_len(pcie_doe_get_req(doe_cap)))
+                            discard = doe->protocols[p].set_rsp(doe_cap);
+                        break;
+                    }
+                }
+
+                /* Set DOE Ready */
+                if (!discard) {
+                    pcie_doe_set_ready(doe_cap, 1);
+                } else {
+                    pcie_doe_reset_mbox(doe_cap);
+                }
+            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_INTR_EN)) {
+                doe_cap->ctrl.intr = 1;
+            }
+            break;
+        case PCI_EXP_DOE_RD_DATA_MBOX:
+            /* Read MBOX */
+            DOE_DBG("  INCR RD MBOX DO DW OFFSET=%d\n", doe_cap->read_mbox_idx);
+            doe_cap->read_mbox_idx += 1;
+            /* Last DW */
+            if (doe_cap->read_mbox_idx >= doe_cap->read_mbox_len) {
+                pcie_doe_reset_mbox(doe_cap);
+                pcie_doe_set_ready(doe_cap, 0);
+            }
+            break;
+        case PCI_EXP_DOE_WR_DATA_MBOX:
+            /* Write MBOX */
+            DOE_DBG("  WR MBOX=%x, DW OFFSET = %d\n", val, doe_cap->write_mbox_len);
+            pcie_doe_write_mbox(doe_cap, val);
+            break;
+        case PCI_EXP_DOE_STATUS:
+        case PCI_EXP_DOE_CAP:
+        default:
+            break;
+        }
+    }
+}
diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
index 76bf3ed..636b2e8 100644
--- a/include/hw/pci/pci_ids.h
+++ b/include/hw/pci/pci_ids.h
@@ -157,6 +157,8 @@
 
 /* Vendors and devices.  Sort key: vendor first, device next. */
 
+#define PCI_VENDOR_ID_PCI_SIG            0x0001
+
 #define PCI_VENDOR_ID_LSI_LOGIC          0x1000
 #define PCI_DEVICE_ID_LSI_53C810         0x0001
 #define PCI_DEVICE_ID_LSI_53C895A        0x0012
diff --git a/include/hw/pci/pcie.h b/include/hw/pci/pcie.h
index 14c58eb..47d6f66 100644
--- a/include/hw/pci/pcie.h
+++ b/include/hw/pci/pcie.h
@@ -25,6 +25,7 @@
 #include "hw/pci/pcie_regs.h"
 #include "hw/pci/pcie_aer.h"
 #include "hw/hotplug.h"
+#include "hw/pci/pcie_doe.h"
 
 typedef enum {
     /* for attention and power indicator */
diff --git a/include/hw/pci/pcie_doe.h b/include/hw/pci/pcie_doe.h
new file mode 100644
index 0000000..f497075
--- /dev/null
+++ b/include/hw/pci/pcie_doe.h
@@ -0,0 +1,166 @@
+#ifndef PCIE_DOE_H
+#define PCIE_DOE_H
+
+#include "qemu/range.h"
+#include "qemu/typedefs.h"
+#include "hw/register.h"
+
+/* 
+ * PCI DOE register defines 7.9.xx 
+ */
+/* DOE Capabilities Register */
+#define PCI_EXP_DOE_CAP             0x04
+REG32(PCI_DOE_CAP_REG, 0)
+    FIELD(PCI_DOE_CAP_REG, INTR_SUPP, 0, 1)
+    FIELD(PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM, 1, 11)
+
+/* DOE Control Register */
+#define PCI_EXP_DOE_CTRL            0x08
+REG32(PCI_DOE_CAP_CONTROL, 0)
+    FIELD(PCI_DOE_CAP_CONTROL, DOE_ABORT, 0, 1)
+    FIELD(PCI_DOE_CAP_CONTROL, DOE_INTR_EN, 1, 1)
+    FIELD(PCI_DOE_CAP_CONTROL, DOE_GO, 31, 1)
+
+/* DOE Status Register  */
+#define PCI_EXP_DOE_STATUS          0x0c
+REG32(PCI_DOE_CAP_STATUS, 0)
+    FIELD(PCI_DOE_CAP_STATUS, DOE_BUSY, 0, 1)
+    FIELD(PCI_DOE_CAP_STATUS, DOE_INTR_STATUS, 1, 1)
+    FIELD(PCI_DOE_CAP_STATUS, DOE_ERROR, 2, 1)
+    FIELD(PCI_DOE_CAP_STATUS, DATA_OBJ_RDY, 31, 1)
+
+/* DOE Write Data Mailbox Register  */
+#define PCI_EXP_DOE_WR_DATA_MBOX    0x10
+
+/* DOE Read Data Mailbox Register  */
+#define PCI_EXP_DOE_RD_DATA_MBOX    0x14
+
+/* 
+ * PCI DOE register defines 7.9.xx 
+ */
+/* Table 7-x2 */
+#define PCI_SIG_DOE_DISCOVERY       0x00
+#define PCI_SIG_DOE_CMA             0x01
+
+#define DATA_OBJ_BUILD_HEADER1(v, p)  ((p << 16) | v)
+
+/* Figure 6-x1 */
+#define DATA_OBJECT_HEADER1_OFFSET  0
+#define DATA_OBJECT_HEADER2_OFFSET  1
+#define DATA_OBJECT_CONTENT_OFFSET  2
+
+#define PCI_DOE_MAX_DW_SIZE (1 << 18)
+#define PCI_DOE_PROTOCOL_MAX 256
+
+/*
+ * Self-defined Marco
+ */
+/* Request/Response State */
+#define DOE_SUCCESS 0
+#define DOE_DISCARD 1
+
+/* An invalid vector number for MSI/MSI-x */
+#define DOE_NO_INTR (~0)
+
+/*
+ * DOE Protocol - Data Object
+ */
+typedef struct DOEHeader DOEHeader;
+typedef struct DOEProtocol DOEProtocol;
+typedef struct PCIEDOE PCIEDOE;
+typedef struct DOECap DOECap;
+
+struct DOEHeader {
+    uint16_t vendor_id;
+    uint8_t doe_type;
+    uint8_t reserved;
+    struct {
+        uint32_t length:18;
+        uint32_t reserved2:14;
+    };
+} QEMU_PACKED;
+
+/* Protocol infos and rsp function callback */
+struct DOEProtocol {
+    uint16_t vendor_id;
+    uint8_t doe_type;
+
+    bool (*set_rsp)(DOECap *);
+};
+
+struct PCIEDOE {
+    PCIDevice *pdev;
+    DOECap *head;
+
+    /* Protocols and its callback response */
+    DOEProtocol protocols[PCI_DOE_PROTOCOL_MAX];
+    uint32_t protocol_num;
+};
+
+struct DOECap {
+    PCIEDOE *doe;
+
+    /* Capability offset */
+    uint16_t offset;
+
+    /* Next DOE capability */
+    DOECap *next;
+
+    /* Capability */
+    struct {
+        bool intr;
+        uint16_t vec;
+    } cap;
+
+    /* Control */
+    struct {
+        bool abort;
+        bool intr;
+        bool go;
+    } ctrl;
+
+    /* Status */
+    struct {
+        bool busy;
+        bool intr;
+        bool error;
+        bool ready;
+    } status;
+
+    /* Mailbox buffer for device */
+    uint32_t *write_mbox;
+    uint32_t *read_mbox;
+
+    /* Mailbox position indicator */
+    uint32_t read_mbox_idx;
+    uint32_t read_mbox_len;
+    uint32_t write_mbox_len;
+};
+
+void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe);
+int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec);
+void pcie_doe_uninit(PCIEDOE *doe);
+void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
+                                uint8_t doe_type,
+                                bool (*set_rsp)(DOECap *));
+uint32_t pcie_doe_build_protocol(DOEProtocol *p);
+DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr);
+uint32_t pcie_doe_read_config(DOECap *doe_cap, uint32_t addr, int size);
+void pcie_doe_write_config(DOECap *doe_cap, uint32_t addr,
+                           uint32_t val, int size);
+
+/* Utility functions for set_rsp in DOEProtocol */
+void *pcie_doe_get_req(DOECap *doe_cap);
+void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp);
+uint32_t pcie_doe_object_len(void *obj);
+
+#define DOE_DEBUG
+#ifdef DOE_DEBUG
+#define DOE_DBG(...) printf(__VA_ARGS__)
+#else
+#define DOE_DBG(...) {}
+#endif
+
+#define dwsizeof(s) ((sizeof(s) + 4 - 1) / 4)
+
+#endif /* PCIE_DOE_H */
diff --git a/include/hw/pci/pcie_regs.h b/include/hw/pci/pcie_regs.h
index 1db86b0..963dc2e 100644
--- a/include/hw/pci/pcie_regs.h
+++ b/include/hw/pci/pcie_regs.h
@@ -179,4 +179,8 @@ typedef enum PCIExpLinkWidth {
 #define PCI_ACS_VER                     0x1
 #define PCI_ACS_SIZEOF                  8
 
+/* DOE Capability Register Fields */
+#define PCI_DOE_VER                     0x1
+#define PCI_DOE_SIZEOF                  24
+
 #endif /* QEMU_PCIE_REGS_H */
diff --git a/include/standard-headers/linux/pci_regs.h b/include/standard-headers/linux/pci_regs.h
index e709ae8..4a7b7a4 100644
--- a/include/standard-headers/linux/pci_regs.h
+++ b/include/standard-headers/linux/pci_regs.h
@@ -730,7 +730,8 @@
 #define PCI_EXT_CAP_ID_DVSEC	0x23	/* Designated Vendor-Specific */
 #define PCI_EXT_CAP_ID_DLF	0x25	/* Data Link Feature */
 #define PCI_EXT_CAP_ID_PL_16GT	0x26	/* Physical Layer 16.0 GT/s */
-#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_PL_16GT
+#define PCI_EXT_CAP_ID_DOE      0x2E    /* Data Object Exchange */
+#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_DOE
 
 #define PCI_EXT_CAP_DSN_SIZEOF	12
 #define PCI_EXT_CAP_MCAST_ENDPOINT_SIZEOF 40
-- 
1.8.3.1

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Ben Widawsky]

Have you/Jonathan come to consensus about which implementation is going forward?
I'd rather not have to review two :D

On 21-02-09 15:35:49, Chris Browy wrote:
> ---
>  MAINTAINERS                               |   7 +
>  hw/pci/meson.build                        |   1 +
>  hw/pci/pcie.c                             |   2 +-
>  hw/pci/pcie_doe.c                         | 414 ++++++++++++++++++++++++++++++
>  include/hw/pci/pci_ids.h                  |   2 +
>  include/hw/pci/pcie.h                     |   1 +
>  include/hw/pci/pcie_doe.h                 | 166 ++++++++++++
>  include/hw/pci/pcie_regs.h                |   4 +
>  include/standard-headers/linux/pci_regs.h |   3 +-
>  9 files changed, 598 insertions(+), 2 deletions(-)
>  create mode 100644 hw/pci/pcie_doe.c
>  create mode 100644 include/hw/pci/pcie_doe.h
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 981dc92..4fb865e 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -1655,6 +1655,13 @@ F: docs/pci*
>  F: docs/specs/*pci*
>  F: default-configs/pci.mak
>  
> +PCIE DOE
> +M: Huai-Cheng Kuo <hchkuo@avery-design.com.tw>
> +M: Chris Browy <cbrowy@avery-design.com>
> +S: Supported
> +F: include/hw/pci/pcie_doe.h
> +F: hw/pci/pcie_doe.c
> +
>  ACPI/SMBIOS
>  M: Michael S. Tsirkin <mst@redhat.com>
>  M: Igor Mammedov <imammedo@redhat.com>
> diff --git a/hw/pci/meson.build b/hw/pci/meson.build
> index 5c4bbac..115e502 100644
> --- a/hw/pci/meson.build
> +++ b/hw/pci/meson.build
> @@ -12,6 +12,7 @@ pci_ss.add(files(
>  # allow plugging PCIe devices into PCI buses, include them even if
>  # CONFIG_PCI_EXPRESS=n.
>  pci_ss.add(files('pcie.c', 'pcie_aer.c'))
> +pci_ss.add(files('pcie_doe.c'))

It looks like this should be like the below line:
softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: pci_doe.c))

>  softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: files('pcie_port.c', 'pcie_host.c'))
>  softmmu_ss.add_all(when: 'CONFIG_PCI', if_true: pci_ss)
>  
> diff --git a/hw/pci/pcie.c b/hw/pci/pcie.c
> index 1ecf6f6..f7516c4 100644
> --- a/hw/pci/pcie.c
> +++ b/hw/pci/pcie.c
> @@ -735,7 +735,7 @@ void pcie_cap_slot_write_config(PCIDevice *dev,
>  
>      hotplug_event_notify(dev);
>  
> -    /* 
> +    /*

Please drop this.

>       * 6.7.3.2 Command Completed Events
>       *
>       * Software issues a command to a hot-plug capable Downstream Port by
> diff --git a/hw/pci/pcie_doe.c b/hw/pci/pcie_doe.c
> new file mode 100644
> index 0000000..df8e92e
> --- /dev/null
> +++ b/hw/pci/pcie_doe.c
> @@ -0,0 +1,414 @@
> +#include "qemu/osdep.h"
> +#include "qemu/log.h"
> +#include "qemu/error-report.h"
> +#include "qapi/error.h"
> +#include "qemu/range.h"
> +#include "hw/pci/pci.h"
> +#include "hw/pci/pcie.h"
> +#include "hw/pci/pcie_doe.h"
> +#include "hw/pci/msi.h"
> +#include "hw/pci/msix.h"
> +
> +/*
> + * DOE Default Protocols (Discovery, CMA)
> + */
> +/* Discovery Request Object */
> +struct doe_discovery {
> +    DOEHeader header;
> +    uint8_t index;
> +    uint8_t reserved[3];
> +} QEMU_PACKED;
> +
> +/* Discovery Response Object */
> +struct doe_discovery_rsp {
> +    DOEHeader header;
> +    uint16_t vendor_id;
> +    uint8_t doe_type;
> +    uint8_t next_index;
> +} QEMU_PACKED;
> +
> +/* Callback for Discovery */
> +static bool pcie_doe_discovery_rsp(DOECap *doe_cap)
> +{
> +    PCIEDOE *doe = doe_cap->doe;
> +    struct doe_discovery *req = pcie_doe_get_req(doe_cap);
> +    uint8_t index = req->index;
> +    DOEProtocol *prot = NULL;
> +
> +    /* Request length mismatch, discard */
> +    if (req->header.length < dwsizeof(struct doe_discovery)) {

Use DIV_ROUND_UP instead of rolling your own thing.

> +        return DOE_DISCARD;
> +    }
> +
> +    /* Point to the requested protocol */
> +    if (index < doe->protocol_num) {
> +        prot = &doe->protocols[index];
> +    }

What happens on else, should that still return DOE_SUCCESS?

> +
> +    struct doe_discovery_rsp rsp = {
> +        .header = {
> +            .vendor_id = PCI_VENDOR_ID_PCI_SIG,
> +            .doe_type = PCI_SIG_DOE_DISCOVERY,
> +            .reserved = 0x0,
> +            .length = dwsizeof(struct doe_discovery_rsp),
> +        },

mixed declarations are not allowed.
DIV_ROUND_UP

> +        .vendor_id = (prot) ? prot->vendor_id : 0xFFFF,
> +        .doe_type = (prot) ? prot->doe_type : 0xFF,
> +        .next_index = (index + 1) < doe->protocol_num ?
> +                      (index + 1) : 0,
> +    };

I prefer:
next_index = (index + 1) % doe->protocol_num

> +
> +    pcie_doe_set_rsp(doe_cap, &rsp);
> +
> +    return DOE_SUCCESS;
> +}
> +
> +/* Callback for CMA */
> +static bool pcie_doe_cma_rsp(DOECap *doe_cap)
> +{
> +    doe_cap->status.error = 1;
> +
> +    memset(doe_cap->read_mbox, 0,
> +           PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +
> +    doe_cap->write_mbox_len = 0;
> +
> +    return DOE_DISCARD;
> +}
> +
> +/*
> + * DOE Utilities
> + */
> +static void pcie_doe_reset_mbox(DOECap *st)
> +{
> +    st->read_mbox_idx = 0;
> +
> +    st->read_mbox_len = 0;
> +    st->write_mbox_len = 0;
> +
> +    memset(st->read_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +    memset(st->write_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +}
> +
> +/*
> + * Initialize the list and protocol for a device.
> + * This function won't add the DOE capabitity to your PCIe device.
> + */
> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe)
> +{
> +    doe->pdev = dev;
> +    doe->head = NULL;
> +    doe->protocol_num = 0;
> +
> +    /* Register two default protocol */
> +    //TODO : LINK LIST

Please do this for next rev of the patches.

> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
> +            PCI_SIG_DOE_DISCOVERY, pcie_doe_discovery_rsp);
> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
> +            PCI_SIG_DOE_CMA, pcie_doe_cma_rsp);
> +}
> +
> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec) {
> +    DOECap *new_cap, **ptr;
> +    PCIDevice *dev = doe->pdev;
> +
> +    pcie_add_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_VER, offset,
> +                        PCI_DOE_SIZEOF);
> +
> +    ptr = &doe->head;
> +    /* Insert the new DOE at the end of linked list */
> +    while (*ptr) {
> +        if (range_covers_byte((*ptr)->offset, PCI_DOE_SIZEOF, offset) ||
> +            (*ptr)->cap.vec == vec) {
> +            return -EINVAL;
> +        }
> +
> +        ptr = &(*ptr)->next;
> +    }
> +    new_cap = g_malloc0(sizeof(DOECap));
> +    *ptr = new_cap;
> +
> +    new_cap->doe = doe;
> +    new_cap->offset = offset;
> +    new_cap->next = NULL;
> +
> +    /* Configure MSI/MSI-X */
> +    if (intr && (msi_present(dev) || msix_present(dev))) {
> +        new_cap->cap.intr = intr;
> +        new_cap->cap.vec = vec;
> +    }
> +
> +    /* Set up W/R Mailbox buffer */
> +    new_cap->write_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +    new_cap->read_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +
> +    pcie_doe_reset_mbox(new_cap);
> +
> +    return 0;
> +}
> +
> +void pcie_doe_uninit(PCIEDOE *doe) {

fini is the more idiomatically unix name for de/un init.
> +    PCIDevice *dev = doe->pdev;
> +    DOECap *cap;
> +
> +    pci_del_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_SIZEOF);
> +
> +    cap = doe->head;
> +    while (cap) {
> +        doe->head = doe->head->next;
> +
> +        g_free(cap->read_mbox);
> +        g_free(cap->write_mbox);
> +        cap->read_mbox = NULL;
> +        cap->write_mbox = NULL;
> +        g_free(cap);
> +        cap = doe->head;
> +    }
> +}
> +
> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
> +        uint8_t doe_type, bool (*set_rsp)(DOECap *))
> +{
> +    /* Protocol num should not exceed 256 */
> +    assert(doe->protocol_num < PCI_DOE_PROTOCOL_MAX);
> +
> +    doe->protocols[doe->protocol_num].vendor_id = vendor_id;
> +    doe->protocols[doe->protocol_num].doe_type = doe_type;
> +    doe->protocols[doe->protocol_num].set_rsp = set_rsp;
> +
> +    doe->protocol_num++;
> +}
> +
> +uint32_t pcie_doe_build_protocol(DOEProtocol *p)
> +{
> +    return DATA_OBJ_BUILD_HEADER1(p->vendor_id, p->doe_type);
> +}
> +
> +/* Return the pointer of DOE request in write mailbox buffer */
> +void *pcie_doe_get_req(DOECap *doe_cap)
> +{
> +    return doe_cap->write_mbox;
> +}
> +
> +/* Copy the response to read mailbox buffer */
> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp)
> +{
> +    uint32_t len = pcie_doe_object_len(rsp);
> +
> +    memcpy(doe_cap->read_mbox + doe_cap->read_mbox_len,
> +           rsp, len * sizeof(uint32_t));
> +    doe_cap->read_mbox_len += len;
> +}
> +
> +/* Get Data Object length */
> +uint32_t pcie_doe_object_len(void *obj)
> +{
> +    return (obj)? ((DOEHeader *)obj)->length : 0;
> +}
> +
> +static void pcie_doe_write_mbox(DOECap *doe_cap, uint32_t val)
> +{
> +    memcpy(doe_cap->write_mbox + doe_cap->write_mbox_len, &val, sizeof(uint32_t));

doe_cap->write_mbox[doe_cap->write_mbox_len] = val?

> +
> +    if (doe_cap->write_mbox_len == 1) {
> +        DOE_DBG("  Capture DOE request DO length = %d\n", val & 0x0003ffff);
> +    }

I don't have the spec in front of me, but this is begging for a comment on why 1
is special.

> +
> +    doe_cap->write_mbox_len++;
> +}
> +
> +static void pcie_doe_irq_assert(DOECap *doe_cap)
> +{
> +    PCIDevice *dev = doe_cap->doe->pdev;
> +
> +    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {
> +        /* Interrupt notify */
> +        if (msix_enabled(dev)) {
> +            msix_notify(dev, doe_cap->cap.vec);
> +        } else if (msi_enabled(dev)) {
> +            msi_notify(dev, doe_cap->cap.vec);
> +        }
> +        /* Not support legacy IRQ */
> +    }
> +}
> +
> +static void pcie_doe_set_ready(DOECap *doe_cap, bool rdy)
> +{
> +    doe_cap->status.ready = rdy;
> +
> +    if (rdy) {
> +        pcie_doe_irq_assert(doe_cap);
> +    }
> +}
> +
> +static void pcie_doe_set_error(DOECap *doe_cap, bool err)
> +{
> +    doe_cap->status.error = err;
> +
> +    if (err) {
> +        pcie_doe_irq_assert(doe_cap);
> +    }
> +}
> +
> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr)
> +{
> +    DOECap *ptr = doe->head;
> +
> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
> +    while (ptr && 
> +           !range_covers_byte(ptr->offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
> +        ptr = ptr->next;
> +    }
> +    
> +    return ptr;
> +}
> +
> +uint32_t pcie_doe_read_config(DOECap *doe_cap,
> +                              uint32_t addr, int size)
> +{
> +    uint32_t ret = 0, shift, mask = 0xFFFFFFFF;
> +    uint16_t doe_offset = doe_cap->offset;
> +
> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
> +        addr -= doe_offset;
> +
> +        if (range_covers_byte(PCI_EXP_DOE_CAP, sizeof(uint32_t), addr)) {
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, INTR_SUPP,
> +                             doe_cap->cap.intr);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM,
> +                             doe_cap->cap.vec);
> +        } else if (range_covers_byte(PCI_EXP_DOE_CTRL, sizeof(uint32_t), addr)) {
> +            /* Must return ABORT=0 and GO=0 */
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_CONTROL, DOE_INTR_EN,
> +                             doe_cap->ctrl.intr);
> +            DOE_DBG("  CONTROL REG=%x\n", ret);
> +        } else if (range_covers_byte(PCI_EXP_DOE_STATUS, sizeof(uint32_t), addr)) {
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_BUSY,
> +                             doe_cap->status.busy);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_INTR_STATUS,
> +                             doe_cap->status.intr);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_ERROR,
> +                             doe_cap->status.error);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DATA_OBJ_RDY,
> +                             doe_cap->status.ready);
> +        } else if (range_covers_byte(PCI_EXP_DOE_RD_DATA_MBOX, sizeof(uint32_t), addr)) {
> +            /* Check that READY is set */
> +            if (!doe_cap->status.ready) {
> +                /* Return 0 if not ready */
> +                DOE_DBG("  RD MBOX RETURN=%x\n", ret);
> +            } else {
> +                /* Deposit next DO DW into read mbox */
> +                DOE_DBG("  RD MBOX, DATA OBJECT READY,"
> +                        " CURRENT DO DW OFFSET=%x\n",
> +                        doe_cap->read_mbox_idx);
> +
> +                ret = doe_cap->read_mbox[doe_cap->read_mbox_idx];
> +
> +                DOE_DBG("  RD MBOX DW=%x\n", ret);
> +                DOE_DBG("  RD MBOX DW OFFSET=%d, RD MBOX LENGTH=%d\n",
> +                        doe_cap->read_mbox_idx, doe_cap->read_mbox_len);
> +            }
> +        } else if (range_covers_byte(PCI_EXP_DOE_WR_DATA_MBOX, sizeof(uint32_t), addr)) {
> +            DOE_DBG("  WR MBOX, local_val =%x\n", ret);
> +        }
> +    }
> +
> +    /* Alignment */
> +    shift = addr % sizeof(uint32_t);

Can these actually be unaligned? The whole range_covers_byte() stuff could go
away if not.

> +    if (shift) {
> +        ret >>= shift * 8;
> +    }
> +    mask >>= (sizeof(uint32_t) - size) * 8;
> +
> +    return ret & mask;
> +}
> +
> +void pcie_doe_write_config(DOECap *doe_cap,
> +                           uint32_t addr, uint32_t val, int size)
> +{
> +    PCIEDOE *doe = doe_cap->doe;
> +    uint16_t doe_offset = doe_cap->offset;
> +    int p;
> +    bool discard;
> +    uint32_t shift;
> +
> +    DOE_DBG("  addr=%x, val=%x, size=%x\n", addr, val, size);
> +
> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
> +        /* Alignment */
> +        shift = addr % sizeof(uint32_t);
> +        addr -= (doe_offset - shift);
> +        val <<= shift * 8;
> +
> +        switch (addr) {
> +        case PCI_EXP_DOE_CTRL:
> +            DOE_DBG("  CONTROL=%x\n", val);
> +            if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_ABORT)) {
> +                /* If ABORT, clear status reg DOE Error and DOE Ready */
> +                DOE_DBG("  Setting ABORT\n");
> +                pcie_doe_set_ready(doe_cap, 0);
> +                pcie_doe_set_error(doe_cap, 0);
> +                pcie_doe_reset_mbox(doe_cap);
> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_GO)) {
> +                DOE_DBG("  CONTROL GO=%x\n", val);
> +                /*
> +                 * Check protocol the incoming request in write_mbox and
> +                 * memcpy the corresponding response to read_mbox.
> +                 *
> +                 * "discard" should be set up if the response callback
> +                 * function could not deal with request (e.g. length
> +                 * mismatch) or the protocol of request was not found.
> +                 */
> +                discard = DOE_DISCARD;
> +                for (p = 0; p < doe->protocol_num; p++) {
> +                    if (doe_cap->write_mbox[0] ==
> +                        pcie_doe_build_protocol(&doe->protocols[p])) {
> +                        /* Found */
> +                        DOE_DBG("  DOE PROTOCOL=%x\n", doe_cap->write_mbox[0]);
> +                        /*
> +                         * Spec:
> +                         * If the number of DW transferred does not match the
> +                         * indicated Length for a data object, then the
> +                         * data object must be silently discarded.
> +                         */
> +                        if (doe_cap->write_mbox_len ==
> +                            pcie_doe_object_len(pcie_doe_get_req(doe_cap)))
> +                            discard = doe->protocols[p].set_rsp(doe_cap);
> +                        break;
> +                    }
> +                }
> +
> +                /* Set DOE Ready */
> +                if (!discard) {
> +                    pcie_doe_set_ready(doe_cap, 1);
> +                } else {
> +                    pcie_doe_reset_mbox(doe_cap);
> +                }
> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_INTR_EN)) {
> +                doe_cap->ctrl.intr = 1;
> +            }
> +            break;
> +        case PCI_EXP_DOE_RD_DATA_MBOX:
> +            /* Read MBOX */
> +            DOE_DBG("  INCR RD MBOX DO DW OFFSET=%d\n", doe_cap->read_mbox_idx);
> +            doe_cap->read_mbox_idx += 1;
> +            /* Last DW */
> +            if (doe_cap->read_mbox_idx >= doe_cap->read_mbox_len) {
> +                pcie_doe_reset_mbox(doe_cap);
> +                pcie_doe_set_ready(doe_cap, 0);
> +            }
> +            break;
> +        case PCI_EXP_DOE_WR_DATA_MBOX:
> +            /* Write MBOX */
> +            DOE_DBG("  WR MBOX=%x, DW OFFSET = %d\n", val, doe_cap->write_mbox_len);
> +            pcie_doe_write_mbox(doe_cap, val);
> +            break;
> +        case PCI_EXP_DOE_STATUS:
> +        case PCI_EXP_DOE_CAP:
> +        default:
> +            break;
> +        }
> +    }
> +}
> diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
> index 76bf3ed..636b2e8 100644
> --- a/include/hw/pci/pci_ids.h
> +++ b/include/hw/pci/pci_ids.h
> @@ -157,6 +157,8 @@
>  
>  /* Vendors and devices.  Sort key: vendor first, device next. */
>  
> +#define PCI_VENDOR_ID_PCI_SIG            0x0001
> +
>  #define PCI_VENDOR_ID_LSI_LOGIC          0x1000
>  #define PCI_DEVICE_ID_LSI_53C810         0x0001
>  #define PCI_DEVICE_ID_LSI_53C895A        0x0012
> diff --git a/include/hw/pci/pcie.h b/include/hw/pci/pcie.h
> index 14c58eb..47d6f66 100644
> --- a/include/hw/pci/pcie.h
> +++ b/include/hw/pci/pcie.h
> @@ -25,6 +25,7 @@
>  #include "hw/pci/pcie_regs.h"
>  #include "hw/pci/pcie_aer.h"
>  #include "hw/hotplug.h"
> +#include "hw/pci/pcie_doe.h"
>  
>  typedef enum {
>      /* for attention and power indicator */
> diff --git a/include/hw/pci/pcie_doe.h b/include/hw/pci/pcie_doe.h
> new file mode 100644
> index 0000000..f497075
> --- /dev/null
> +++ b/include/hw/pci/pcie_doe.h
> @@ -0,0 +1,166 @@
> +#ifndef PCIE_DOE_H
> +#define PCIE_DOE_H
> +
> +#include "qemu/range.h"
> +#include "qemu/typedefs.h"
> +#include "hw/register.h"
> +
> +/* 
> + * PCI DOE register defines 7.9.xx 
> + */

Whitespace issues

> +/* DOE Capabilities Register */
> +#define PCI_EXP_DOE_CAP             0x04
> +REG32(PCI_DOE_CAP_REG, 0)
> +    FIELD(PCI_DOE_CAP_REG, INTR_SUPP, 0, 1)
> +    FIELD(PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM, 1, 11)
> +
> +/* DOE Control Register */
> +#define PCI_EXP_DOE_CTRL            0x08
> +REG32(PCI_DOE_CAP_CONTROL, 0)
> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_ABORT, 0, 1)
> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_INTR_EN, 1, 1)
> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_GO, 31, 1)
> +
> +/* DOE Status Register  */
> +#define PCI_EXP_DOE_STATUS          0x0c
> +REG32(PCI_DOE_CAP_STATUS, 0)
> +    FIELD(PCI_DOE_CAP_STATUS, DOE_BUSY, 0, 1)
> +    FIELD(PCI_DOE_CAP_STATUS, DOE_INTR_STATUS, 1, 1)
> +    FIELD(PCI_DOE_CAP_STATUS, DOE_ERROR, 2, 1)
> +    FIELD(PCI_DOE_CAP_STATUS, DATA_OBJ_RDY, 31, 1)
> +
> +/* DOE Write Data Mailbox Register  */
> +#define PCI_EXP_DOE_WR_DATA_MBOX    0x10
> +
> +/* DOE Read Data Mailbox Register  */
> +#define PCI_EXP_DOE_RD_DATA_MBOX    0x14
> +
> +/* 
> + * PCI DOE register defines 7.9.xx 
> + */

Is this duplicated on purpose?

> +/* Table 7-x2 */
> +#define PCI_SIG_DOE_DISCOVERY       0x00
> +#define PCI_SIG_DOE_CMA             0x01
> +
> +#define DATA_OBJ_BUILD_HEADER1(v, p)  ((p << 16) | v)
> +
> +/* Figure 6-x1 */
> +#define DATA_OBJECT_HEADER1_OFFSET  0
> +#define DATA_OBJECT_HEADER2_OFFSET  1
> +#define DATA_OBJECT_CONTENT_OFFSET  2
> +
> +#define PCI_DOE_MAX_DW_SIZE (1 << 18)
> +#define PCI_DOE_PROTOCOL_MAX 256
> +
> +/*
> + * Self-defined Marco
> + */
> +/* Request/Response State */
> +#define DOE_SUCCESS 0
> +#define DOE_DISCARD 1
> +
> +/* An invalid vector number for MSI/MSI-x */
> +#define DOE_NO_INTR (~0)
> +
> +/*
> + * DOE Protocol - Data Object
> + */
> +typedef struct DOEHeader DOEHeader;
> +typedef struct DOEProtocol DOEProtocol;
> +typedef struct PCIEDOE PCIEDOE;
> +typedef struct DOECap DOECap;
> +
> +struct DOEHeader {
> +    uint16_t vendor_id;
> +    uint8_t doe_type;
> +    uint8_t reserved;
> +    struct {
> +        uint32_t length:18;
> +        uint32_t reserved2:14;
> +    };
> +} QEMU_PACKED;
> +
> +/* Protocol infos and rsp function callback */
> +struct DOEProtocol {
> +    uint16_t vendor_id;
> +    uint8_t doe_type;
> +
> +    bool (*set_rsp)(DOECap *);
> +};
> +
> +struct PCIEDOE {
> +    PCIDevice *pdev;
> +    DOECap *head;
> +
> +    /* Protocols and its callback response */
> +    DOEProtocol protocols[PCI_DOE_PROTOCOL_MAX];
> +    uint32_t protocol_num;
> +};
> +
> +struct DOECap {
> +    PCIEDOE *doe;
> +
> +    /* Capability offset */
> +    uint16_t offset;
> +
> +    /* Next DOE capability */
> +    DOECap *next;
> +
> +    /* Capability */
> +    struct {
> +        bool intr;
> +        uint16_t vec;
> +    } cap;
> +
> +    /* Control */
> +    struct {
> +        bool abort;
> +        bool intr;
> +        bool go;
> +    } ctrl;
> +
> +    /* Status */
> +    struct {
> +        bool busy;
> +        bool intr;
> +        bool error;
> +        bool ready;
> +    } status;
> +
> +    /* Mailbox buffer for device */
> +    uint32_t *write_mbox;
> +    uint32_t *read_mbox;
> +
> +    /* Mailbox position indicator */
> +    uint32_t read_mbox_idx;
> +    uint32_t read_mbox_len;
> +    uint32_t write_mbox_len;
> +};
> +
> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe);
> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec);
> +void pcie_doe_uninit(PCIEDOE *doe);
> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
> +                                uint8_t doe_type,
> +                                bool (*set_rsp)(DOECap *));
> +uint32_t pcie_doe_build_protocol(DOEProtocol *p);
> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr);
> +uint32_t pcie_doe_read_config(DOECap *doe_cap, uint32_t addr, int size);
> +void pcie_doe_write_config(DOECap *doe_cap, uint32_t addr,
> +                           uint32_t val, int size);
> +
> +/* Utility functions for set_rsp in DOEProtocol */
> +void *pcie_doe_get_req(DOECap *doe_cap);
> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp);
> +uint32_t pcie_doe_object_len(void *obj);
> +
> +#define DOE_DEBUG
> +#ifdef DOE_DEBUG
> +#define DOE_DBG(...) printf(__VA_ARGS__)
> +#else
> +#define DOE_DBG(...) {}
> +#endif
> +
> +#define dwsizeof(s) ((sizeof(s) + 4 - 1) / 4)
> +
> +#endif /* PCIE_DOE_H */
> diff --git a/include/hw/pci/pcie_regs.h b/include/hw/pci/pcie_regs.h
> index 1db86b0..963dc2e 100644
> --- a/include/hw/pci/pcie_regs.h
> +++ b/include/hw/pci/pcie_regs.h
> @@ -179,4 +179,8 @@ typedef enum PCIExpLinkWidth {
>  #define PCI_ACS_VER                     0x1
>  #define PCI_ACS_SIZEOF                  8
>  > +/* DOE Capability Register Fields */
> +#define PCI_DOE_VER                     0x1
> +#define PCI_DOE_SIZEOF                  24
> +
>  #endif /* QEMU_PCIE_REGS_H */
> diff --git a/include/standard-headers/linux/pci_regs.h b/include/standard-headers/linux/pci_regs.h
> index e709ae8..4a7b7a4 100644
> --- a/include/standard-headers/linux/pci_regs.h
> +++ b/include/standard-headers/linux/pci_regs.h
> @@ -730,7 +730,8 @@
>  #define PCI_EXT_CAP_ID_DVSEC	0x23	/* Designated Vendor-Specific */
>  #define PCI_EXT_CAP_ID_DLF	0x25	/* Data Link Feature */
>  #define PCI_EXT_CAP_ID_PL_16GT	0x26	/* Physical Layer 16.0 GT/s */
> -#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_PL_16GT
> +#define PCI_EXT_CAP_ID_DOE      0x2E    /* Data Object Exchange */
> +#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_DOE
>  
>  #define PCI_EXT_CAP_DSN_SIZEOF	12
>  #define PCI_EXT_CAP_MCAST_ENDPOINT_SIZEOF 40

I haven't reviewed the spec stuff, but I assume Jonathan is familiar with that
and probably knows it almost by heart already.

> -- 
> 1.8.3.1
> 
> 

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Chris Browy]

No consensus yet but I’d suggest that we’ll do the QEMU work and Jonathan focuses 
on the linux kernel and UEFI/edk2 and CXL SSWG efforts.  Seems like
a way to maximize resources and everyone’s contribution and expertise.  QEMU part
requires the least expertise which is why we’re best suited for it compared to other 
areas ;)

Review comments will be folded into next patch.

> On Feb 9, 2021, at 4:42 PM, Ben Widawsky <ben@bwidawsk.net> wrote:
> 
> Have you/Jonathan come to consensus about which implementation is going forward?
> I'd rather not have to review two :D
> 
> On 21-02-09 15:35:49, Chris Browy wrote:
>> ---
>> MAINTAINERS                               |   7 +
>> hw/pci/meson.build                        |   1 +
>> hw/pci/pcie.c                             |   2 +-
>> hw/pci/pcie_doe.c                         | 414 ++++++++++++++++++++++++++++++
>> include/hw/pci/pci_ids.h                  |   2 +
>> include/hw/pci/pcie.h                     |   1 +
>> include/hw/pci/pcie_doe.h                 | 166 ++++++++++++
>> include/hw/pci/pcie_regs.h                |   4 +
>> include/standard-headers/linux/pci_regs.h |   3 +-
>> 9 files changed, 598 insertions(+), 2 deletions(-)
>> create mode 100644 hw/pci/pcie_doe.c
>> create mode 100644 include/hw/pci/pcie_doe.h
>> 
>> diff --git a/MAINTAINERS b/MAINTAINERS
>> index 981dc92..4fb865e 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -1655,6 +1655,13 @@ F: docs/pci*
>> F: docs/specs/*pci*
>> F: default-configs/pci.mak
>> 
>> +PCIE DOE
>> +M: Huai-Cheng Kuo <hchkuo@avery-design.com.tw>
>> +M: Chris Browy <cbrowy@avery-design.com>
>> +S: Supported
>> +F: include/hw/pci/pcie_doe.h
>> +F: hw/pci/pcie_doe.c
>> +
>> ACPI/SMBIOS
>> M: Michael S. Tsirkin <mst@redhat.com>
>> M: Igor Mammedov <imammedo@redhat.com>
>> diff --git a/hw/pci/meson.build b/hw/pci/meson.build
>> index 5c4bbac..115e502 100644
>> --- a/hw/pci/meson.build
>> +++ b/hw/pci/meson.build
>> @@ -12,6 +12,7 @@ pci_ss.add(files(
>> # allow plugging PCIe devices into PCI buses, include them even if
>> # CONFIG_PCI_EXPRESS=n.
>> pci_ss.add(files('pcie.c', 'pcie_aer.c'))
>> +pci_ss.add(files('pcie_doe.c'))
> 
> It looks like this should be like the below line:
> softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: pci_doe.c))
> 
>> softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: files('pcie_port.c', 'pcie_host.c'))
>> softmmu_ss.add_all(when: 'CONFIG_PCI', if_true: pci_ss)
>> 
>> diff --git a/hw/pci/pcie.c b/hw/pci/pcie.c
>> index 1ecf6f6..f7516c4 100644
>> --- a/hw/pci/pcie.c
>> +++ b/hw/pci/pcie.c
>> @@ -735,7 +735,7 @@ void pcie_cap_slot_write_config(PCIDevice *dev,
>> 
>>     hotplug_event_notify(dev);
>> 
>> -    /* 
>> +    /*
> 
> Please drop this.
> 
>>      * 6.7.3.2 Command Completed Events
>>      *
>>      * Software issues a command to a hot-plug capable Downstream Port by
>> diff --git a/hw/pci/pcie_doe.c b/hw/pci/pcie_doe.c
>> new file mode 100644
>> index 0000000..df8e92e
>> --- /dev/null
>> +++ b/hw/pci/pcie_doe.c
>> @@ -0,0 +1,414 @@
>> +#include "qemu/osdep.h"
>> +#include "qemu/log.h"
>> +#include "qemu/error-report.h"
>> +#include "qapi/error.h"
>> +#include "qemu/range.h"
>> +#include "hw/pci/pci.h"
>> +#include "hw/pci/pcie.h"
>> +#include "hw/pci/pcie_doe.h"
>> +#include "hw/pci/msi.h"
>> +#include "hw/pci/msix.h"
>> +
>> +/*
>> + * DOE Default Protocols (Discovery, CMA)
>> + */
>> +/* Discovery Request Object */
>> +struct doe_discovery {
>> +    DOEHeader header;
>> +    uint8_t index;
>> +    uint8_t reserved[3];
>> +} QEMU_PACKED;
>> +
>> +/* Discovery Response Object */
>> +struct doe_discovery_rsp {
>> +    DOEHeader header;
>> +    uint16_t vendor_id;
>> +    uint8_t doe_type;
>> +    uint8_t next_index;
>> +} QEMU_PACKED;
>> +
>> +/* Callback for Discovery */
>> +static bool pcie_doe_discovery_rsp(DOECap *doe_cap)
>> +{
>> +    PCIEDOE *doe = doe_cap->doe;
>> +    struct doe_discovery *req = pcie_doe_get_req(doe_cap);
>> +    uint8_t index = req->index;
>> +    DOEProtocol *prot = NULL;
>> +
>> +    /* Request length mismatch, discard */
>> +    if (req->header.length < dwsizeof(struct doe_discovery)) {
> 
> Use DIV_ROUND_UP instead of rolling your own thing.
> 
>> +        return DOE_DISCARD;
>> +    }
>> +
>> +    /* Point to the requested protocol */
>> +    if (index < doe->protocol_num) {
>> +        prot = &doe->protocols[index];
>> +    }
> 
> What happens on else, should that still return DOE_SUCCESS?
> 
>> +
>> +    struct doe_discovery_rsp rsp = {
>> +        .header = {
>> +            .vendor_id = PCI_VENDOR_ID_PCI_SIG,
>> +            .doe_type = PCI_SIG_DOE_DISCOVERY,
>> +            .reserved = 0x0,
>> +            .length = dwsizeof(struct doe_discovery_rsp),
>> +        },
> 
> mixed declarations are not allowed.
> DIV_ROUND_UP
> 
>> +        .vendor_id = (prot) ? prot->vendor_id : 0xFFFF,
>> +        .doe_type = (prot) ? prot->doe_type : 0xFF,
>> +        .next_index = (index + 1) < doe->protocol_num ?
>> +                      (index + 1) : 0,
>> +    };
> 
> I prefer:
> next_index = (index + 1) % doe->protocol_num
> 
>> +
>> +    pcie_doe_set_rsp(doe_cap, &rsp);
>> +
>> +    return DOE_SUCCESS;
>> +}
>> +
>> +/* Callback for CMA */
>> +static bool pcie_doe_cma_rsp(DOECap *doe_cap)
>> +{
>> +    doe_cap->status.error = 1;
>> +
>> +    memset(doe_cap->read_mbox, 0,
>> +           PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +
>> +    doe_cap->write_mbox_len = 0;
>> +
>> +    return DOE_DISCARD;
>> +}
>> +
>> +/*
>> + * DOE Utilities
>> + */
>> +static void pcie_doe_reset_mbox(DOECap *st)
>> +{
>> +    st->read_mbox_idx = 0;
>> +
>> +    st->read_mbox_len = 0;
>> +    st->write_mbox_len = 0;
>> +
>> +    memset(st->read_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +    memset(st->write_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +}
>> +
>> +/*
>> + * Initialize the list and protocol for a device.
>> + * This function won't add the DOE capabitity to your PCIe device.
>> + */
>> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe)
>> +{
>> +    doe->pdev = dev;
>> +    doe->head = NULL;
>> +    doe->protocol_num = 0;
>> +
>> +    /* Register two default protocol */
>> +    //TODO : LINK LIST
> 
> Please do this for next rev of the patches.
> 
>> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
>> +            PCI_SIG_DOE_DISCOVERY, pcie_doe_discovery_rsp);
>> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
>> +            PCI_SIG_DOE_CMA, pcie_doe_cma_rsp);
>> +}
>> +
>> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec) {
>> +    DOECap *new_cap, **ptr;
>> +    PCIDevice *dev = doe->pdev;
>> +
>> +    pcie_add_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_VER, offset,
>> +                        PCI_DOE_SIZEOF);
>> +
>> +    ptr = &doe->head;
>> +    /* Insert the new DOE at the end of linked list */
>> +    while (*ptr) {
>> +        if (range_covers_byte((*ptr)->offset, PCI_DOE_SIZEOF, offset) ||
>> +            (*ptr)->cap.vec == vec) {
>> +            return -EINVAL;
>> +        }
>> +
>> +        ptr = &(*ptr)->next;
>> +    }
>> +    new_cap = g_malloc0(sizeof(DOECap));
>> +    *ptr = new_cap;
>> +
>> +    new_cap->doe = doe;
>> +    new_cap->offset = offset;
>> +    new_cap->next = NULL;
>> +
>> +    /* Configure MSI/MSI-X */
>> +    if (intr && (msi_present(dev) || msix_present(dev))) {
>> +        new_cap->cap.intr = intr;
>> +        new_cap->cap.vec = vec;
>> +    }
>> +
>> +    /* Set up W/R Mailbox buffer */
>> +    new_cap->write_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +    new_cap->read_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +
>> +    pcie_doe_reset_mbox(new_cap);
>> +
>> +    return 0;
>> +}
>> +
>> +void pcie_doe_uninit(PCIEDOE *doe) {
> 
> fini is the more idiomatically unix name for de/un init.
>> +    PCIDevice *dev = doe->pdev;
>> +    DOECap *cap;
>> +
>> +    pci_del_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_SIZEOF);
>> +
>> +    cap = doe->head;
>> +    while (cap) {
>> +        doe->head = doe->head->next;
>> +
>> +        g_free(cap->read_mbox);
>> +        g_free(cap->write_mbox);
>> +        cap->read_mbox = NULL;
>> +        cap->write_mbox = NULL;
>> +        g_free(cap);
>> +        cap = doe->head;
>> +    }
>> +}
>> +
>> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
>> +        uint8_t doe_type, bool (*set_rsp)(DOECap *))
>> +{
>> +    /* Protocol num should not exceed 256 */
>> +    assert(doe->protocol_num < PCI_DOE_PROTOCOL_MAX);
>> +
>> +    doe->protocols[doe->protocol_num].vendor_id = vendor_id;
>> +    doe->protocols[doe->protocol_num].doe_type = doe_type;
>> +    doe->protocols[doe->protocol_num].set_rsp = set_rsp;
>> +
>> +    doe->protocol_num++;
>> +}
>> +
>> +uint32_t pcie_doe_build_protocol(DOEProtocol *p)
>> +{
>> +    return DATA_OBJ_BUILD_HEADER1(p->vendor_id, p->doe_type);
>> +}
>> +
>> +/* Return the pointer of DOE request in write mailbox buffer */
>> +void *pcie_doe_get_req(DOECap *doe_cap)
>> +{
>> +    return doe_cap->write_mbox;
>> +}
>> +
>> +/* Copy the response to read mailbox buffer */
>> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp)
>> +{
>> +    uint32_t len = pcie_doe_object_len(rsp);
>> +
>> +    memcpy(doe_cap->read_mbox + doe_cap->read_mbox_len,
>> +           rsp, len * sizeof(uint32_t));
>> +    doe_cap->read_mbox_len += len;
>> +}
>> +
>> +/* Get Data Object length */
>> +uint32_t pcie_doe_object_len(void *obj)
>> +{
>> +    return (obj)? ((DOEHeader *)obj)->length : 0;
>> +}
>> +
>> +static void pcie_doe_write_mbox(DOECap *doe_cap, uint32_t val)
>> +{
>> +    memcpy(doe_cap->write_mbox + doe_cap->write_mbox_len, &val, sizeof(uint32_t));
> 
> doe_cap->write_mbox[doe_cap->write_mbox_len] = val?
> 
>> +
>> +    if (doe_cap->write_mbox_len == 1) {
>> +        DOE_DBG("  Capture DOE request DO length = %d\n", val & 0x0003ffff);
>> +    }
> 
> I don't have the spec in front of me, but this is begging for a comment on why 1
> is special.
> 
>> +
>> +    doe_cap->write_mbox_len++;
>> +}
>> +
>> +static void pcie_doe_irq_assert(DOECap *doe_cap)
>> +{
>> +    PCIDevice *dev = doe_cap->doe->pdev;
>> +
>> +    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {
>> +        /* Interrupt notify */
>> +        if (msix_enabled(dev)) {
>> +            msix_notify(dev, doe_cap->cap.vec);
>> +        } else if (msi_enabled(dev)) {
>> +            msi_notify(dev, doe_cap->cap.vec);
>> +        }
>> +        /* Not support legacy IRQ */
>> +    }
>> +}
>> +
>> +static void pcie_doe_set_ready(DOECap *doe_cap, bool rdy)
>> +{
>> +    doe_cap->status.ready = rdy;
>> +
>> +    if (rdy) {
>> +        pcie_doe_irq_assert(doe_cap);
>> +    }
>> +}
>> +
>> +static void pcie_doe_set_error(DOECap *doe_cap, bool err)
>> +{
>> +    doe_cap->status.error = err;
>> +
>> +    if (err) {
>> +        pcie_doe_irq_assert(doe_cap);
>> +    }
>> +}
>> +
>> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr)
>> +{
>> +    DOECap *ptr = doe->head;
>> +
>> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
>> +    while (ptr && 
>> +           !range_covers_byte(ptr->offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
>> +        ptr = ptr->next;
>> +    }
>> +    
>> +    return ptr;
>> +}
>> +
>> +uint32_t pcie_doe_read_config(DOECap *doe_cap,
>> +                              uint32_t addr, int size)
>> +{
>> +    uint32_t ret = 0, shift, mask = 0xFFFFFFFF;
>> +    uint16_t doe_offset = doe_cap->offset;
>> +
>> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
>> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
>> +        addr -= doe_offset;
>> +
>> +        if (range_covers_byte(PCI_EXP_DOE_CAP, sizeof(uint32_t), addr)) {
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, INTR_SUPP,
>> +                             doe_cap->cap.intr);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM,
>> +                             doe_cap->cap.vec);
>> +        } else if (range_covers_byte(PCI_EXP_DOE_CTRL, sizeof(uint32_t), addr)) {
>> +            /* Must return ABORT=0 and GO=0 */
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_CONTROL, DOE_INTR_EN,
>> +                             doe_cap->ctrl.intr);
>> +            DOE_DBG("  CONTROL REG=%x\n", ret);
>> +        } else if (range_covers_byte(PCI_EXP_DOE_STATUS, sizeof(uint32_t), addr)) {
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_BUSY,
>> +                             doe_cap->status.busy);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_INTR_STATUS,
>> +                             doe_cap->status.intr);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_ERROR,
>> +                             doe_cap->status.error);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DATA_OBJ_RDY,
>> +                             doe_cap->status.ready);
>> +        } else if (range_covers_byte(PCI_EXP_DOE_RD_DATA_MBOX, sizeof(uint32_t), addr)) {
>> +            /* Check that READY is set */
>> +            if (!doe_cap->status.ready) {
>> +                /* Return 0 if not ready */
>> +                DOE_DBG("  RD MBOX RETURN=%x\n", ret);
>> +            } else {
>> +                /* Deposit next DO DW into read mbox */
>> +                DOE_DBG("  RD MBOX, DATA OBJECT READY,"
>> +                        " CURRENT DO DW OFFSET=%x\n",
>> +                        doe_cap->read_mbox_idx);
>> +
>> +                ret = doe_cap->read_mbox[doe_cap->read_mbox_idx];
>> +
>> +                DOE_DBG("  RD MBOX DW=%x\n", ret);
>> +                DOE_DBG("  RD MBOX DW OFFSET=%d, RD MBOX LENGTH=%d\n",
>> +                        doe_cap->read_mbox_idx, doe_cap->read_mbox_len);
>> +            }
>> +        } else if (range_covers_byte(PCI_EXP_DOE_WR_DATA_MBOX, sizeof(uint32_t), addr)) {
>> +            DOE_DBG("  WR MBOX, local_val =%x\n", ret);
>> +        }
>> +    }
>> +
>> +    /* Alignment */
>> +    shift = addr % sizeof(uint32_t);
> 
> Can these actually be unaligned? The whole range_covers_byte() stuff could go
> away if not.
> 
>> +    if (shift) {
>> +        ret >>= shift * 8;
>> +    }
>> +    mask >>= (sizeof(uint32_t) - size) * 8;
>> +
>> +    return ret & mask;
>> +}
>> +
>> +void pcie_doe_write_config(DOECap *doe_cap,
>> +                           uint32_t addr, uint32_t val, int size)
>> +{
>> +    PCIEDOE *doe = doe_cap->doe;
>> +    uint16_t doe_offset = doe_cap->offset;
>> +    int p;
>> +    bool discard;
>> +    uint32_t shift;
>> +
>> +    DOE_DBG("  addr=%x, val=%x, size=%x\n", addr, val, size);
>> +
>> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
>> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
>> +        /* Alignment */
>> +        shift = addr % sizeof(uint32_t);
>> +        addr -= (doe_offset - shift);
>> +        val <<= shift * 8;
>> +
>> +        switch (addr) {
>> +        case PCI_EXP_DOE_CTRL:
>> +            DOE_DBG("  CONTROL=%x\n", val);
>> +            if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_ABORT)) {
>> +                /* If ABORT, clear status reg DOE Error and DOE Ready */
>> +                DOE_DBG("  Setting ABORT\n");
>> +                pcie_doe_set_ready(doe_cap, 0);
>> +                pcie_doe_set_error(doe_cap, 0);
>> +                pcie_doe_reset_mbox(doe_cap);
>> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_GO)) {
>> +                DOE_DBG("  CONTROL GO=%x\n", val);
>> +                /*
>> +                 * Check protocol the incoming request in write_mbox and
>> +                 * memcpy the corresponding response to read_mbox.
>> +                 *
>> +                 * "discard" should be set up if the response callback
>> +                 * function could not deal with request (e.g. length
>> +                 * mismatch) or the protocol of request was not found.
>> +                 */
>> +                discard = DOE_DISCARD;
>> +                for (p = 0; p < doe->protocol_num; p++) {
>> +                    if (doe_cap->write_mbox[0] ==
>> +                        pcie_doe_build_protocol(&doe->protocols[p])) {
>> +                        /* Found */
>> +                        DOE_DBG("  DOE PROTOCOL=%x\n", doe_cap->write_mbox[0]);
>> +                        /*
>> +                         * Spec:
>> +                         * If the number of DW transferred does not match the
>> +                         * indicated Length for a data object, then the
>> +                         * data object must be silently discarded.
>> +                         */
>> +                        if (doe_cap->write_mbox_len ==
>> +                            pcie_doe_object_len(pcie_doe_get_req(doe_cap)))
>> +                            discard = doe->protocols[p].set_rsp(doe_cap);
>> +                        break;
>> +                    }
>> +                }
>> +
>> +                /* Set DOE Ready */
>> +                if (!discard) {
>> +                    pcie_doe_set_ready(doe_cap, 1);
>> +                } else {
>> +                    pcie_doe_reset_mbox(doe_cap);
>> +                }
>> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_INTR_EN)) {
>> +                doe_cap->ctrl.intr = 1;
>> +            }
>> +            break;
>> +        case PCI_EXP_DOE_RD_DATA_MBOX:
>> +            /* Read MBOX */
>> +            DOE_DBG("  INCR RD MBOX DO DW OFFSET=%d\n", doe_cap->read_mbox_idx);
>> +            doe_cap->read_mbox_idx += 1;
>> +            /* Last DW */
>> +            if (doe_cap->read_mbox_idx >= doe_cap->read_mbox_len) {
>> +                pcie_doe_reset_mbox(doe_cap);
>> +                pcie_doe_set_ready(doe_cap, 0);
>> +            }
>> +            break;
>> +        case PCI_EXP_DOE_WR_DATA_MBOX:
>> +            /* Write MBOX */
>> +            DOE_DBG("  WR MBOX=%x, DW OFFSET = %d\n", val, doe_cap->write_mbox_len);
>> +            pcie_doe_write_mbox(doe_cap, val);
>> +            break;
>> +        case PCI_EXP_DOE_STATUS:
>> +        case PCI_EXP_DOE_CAP:
>> +        default:
>> +            break;
>> +        }
>> +    }
>> +}
>> diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
>> index 76bf3ed..636b2e8 100644
>> --- a/include/hw/pci/pci_ids.h
>> +++ b/include/hw/pci/pci_ids.h
>> @@ -157,6 +157,8 @@
>> 
>> /* Vendors and devices.  Sort key: vendor first, device next. */
>> 
>> +#define PCI_VENDOR_ID_PCI_SIG            0x0001
>> +
>> #define PCI_VENDOR_ID_LSI_LOGIC          0x1000
>> #define PCI_DEVICE_ID_LSI_53C810         0x0001
>> #define PCI_DEVICE_ID_LSI_53C895A        0x0012
>> diff --git a/include/hw/pci/pcie.h b/include/hw/pci/pcie.h
>> index 14c58eb..47d6f66 100644
>> --- a/include/hw/pci/pcie.h
>> +++ b/include/hw/pci/pcie.h
>> @@ -25,6 +25,7 @@
>> #include "hw/pci/pcie_regs.h"
>> #include "hw/pci/pcie_aer.h"
>> #include "hw/hotplug.h"
>> +#include "hw/pci/pcie_doe.h"
>> 
>> typedef enum {
>>     /* for attention and power indicator */
>> diff --git a/include/hw/pci/pcie_doe.h b/include/hw/pci/pcie_doe.h
>> new file mode 100644
>> index 0000000..f497075
>> --- /dev/null
>> +++ b/include/hw/pci/pcie_doe.h
>> @@ -0,0 +1,166 @@
>> +#ifndef PCIE_DOE_H
>> +#define PCIE_DOE_H
>> +
>> +#include "qemu/range.h"
>> +#include "qemu/typedefs.h"
>> +#include "hw/register.h"
>> +
>> +/* 
>> + * PCI DOE register defines 7.9.xx 
>> + */
> 
> Whitespace issues
> 
>> +/* DOE Capabilities Register */
>> +#define PCI_EXP_DOE_CAP             0x04
>> +REG32(PCI_DOE_CAP_REG, 0)
>> +    FIELD(PCI_DOE_CAP_REG, INTR_SUPP, 0, 1)
>> +    FIELD(PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM, 1, 11)
>> +
>> +/* DOE Control Register */
>> +#define PCI_EXP_DOE_CTRL            0x08
>> +REG32(PCI_DOE_CAP_CONTROL, 0)
>> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_ABORT, 0, 1)
>> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_INTR_EN, 1, 1)
>> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_GO, 31, 1)
>> +
>> +/* DOE Status Register  */
>> +#define PCI_EXP_DOE_STATUS          0x0c
>> +REG32(PCI_DOE_CAP_STATUS, 0)
>> +    FIELD(PCI_DOE_CAP_STATUS, DOE_BUSY, 0, 1)
>> +    FIELD(PCI_DOE_CAP_STATUS, DOE_INTR_STATUS, 1, 1)
>> +    FIELD(PCI_DOE_CAP_STATUS, DOE_ERROR, 2, 1)
>> +    FIELD(PCI_DOE_CAP_STATUS, DATA_OBJ_RDY, 31, 1)
>> +
>> +/* DOE Write Data Mailbox Register  */
>> +#define PCI_EXP_DOE_WR_DATA_MBOX    0x10
>> +
>> +/* DOE Read Data Mailbox Register  */
>> +#define PCI_EXP_DOE_RD_DATA_MBOX    0x14
>> +
>> +/* 
>> + * PCI DOE register defines 7.9.xx 
>> + */
> 
> Is this duplicated on purpose?
> 
>> +/* Table 7-x2 */
>> +#define PCI_SIG_DOE_DISCOVERY       0x00
>> +#define PCI_SIG_DOE_CMA             0x01
>> +
>> +#define DATA_OBJ_BUILD_HEADER1(v, p)  ((p << 16) | v)
>> +
>> +/* Figure 6-x1 */
>> +#define DATA_OBJECT_HEADER1_OFFSET  0
>> +#define DATA_OBJECT_HEADER2_OFFSET  1
>> +#define DATA_OBJECT_CONTENT_OFFSET  2
>> +
>> +#define PCI_DOE_MAX_DW_SIZE (1 << 18)
>> +#define PCI_DOE_PROTOCOL_MAX 256
>> +
>> +/*
>> + * Self-defined Marco
>> + */
>> +/* Request/Response State */
>> +#define DOE_SUCCESS 0
>> +#define DOE_DISCARD 1
>> +
>> +/* An invalid vector number for MSI/MSI-x */
>> +#define DOE_NO_INTR (~0)
>> +
>> +/*
>> + * DOE Protocol - Data Object
>> + */
>> +typedef struct DOEHeader DOEHeader;
>> +typedef struct DOEProtocol DOEProtocol;
>> +typedef struct PCIEDOE PCIEDOE;
>> +typedef struct DOECap DOECap;
>> +
>> +struct DOEHeader {
>> +    uint16_t vendor_id;
>> +    uint8_t doe_type;
>> +    uint8_t reserved;
>> +    struct {
>> +        uint32_t length:18;
>> +        uint32_t reserved2:14;
>> +    };
>> +} QEMU_PACKED;
>> +
>> +/* Protocol infos and rsp function callback */
>> +struct DOEProtocol {
>> +    uint16_t vendor_id;
>> +    uint8_t doe_type;
>> +
>> +    bool (*set_rsp)(DOECap *);
>> +};
>> +
>> +struct PCIEDOE {
>> +    PCIDevice *pdev;
>> +    DOECap *head;
>> +
>> +    /* Protocols and its callback response */
>> +    DOEProtocol protocols[PCI_DOE_PROTOCOL_MAX];
>> +    uint32_t protocol_num;
>> +};
>> +
>> +struct DOECap {
>> +    PCIEDOE *doe;
>> +
>> +    /* Capability offset */
>> +    uint16_t offset;
>> +
>> +    /* Next DOE capability */
>> +    DOECap *next;
>> +
>> +    /* Capability */
>> +    struct {
>> +        bool intr;
>> +        uint16_t vec;
>> +    } cap;
>> +
>> +    /* Control */
>> +    struct {
>> +        bool abort;
>> +        bool intr;
>> +        bool go;
>> +    } ctrl;
>> +
>> +    /* Status */
>> +    struct {
>> +        bool busy;
>> +        bool intr;
>> +        bool error;
>> +        bool ready;
>> +    } status;
>> +
>> +    /* Mailbox buffer for device */
>> +    uint32_t *write_mbox;
>> +    uint32_t *read_mbox;
>> +
>> +    /* Mailbox position indicator */
>> +    uint32_t read_mbox_idx;
>> +    uint32_t read_mbox_len;
>> +    uint32_t write_mbox_len;
>> +};
>> +
>> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe);
>> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec);
>> +void pcie_doe_uninit(PCIEDOE *doe);
>> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
>> +                                uint8_t doe_type,
>> +                                bool (*set_rsp)(DOECap *));
>> +uint32_t pcie_doe_build_protocol(DOEProtocol *p);
>> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr);
>> +uint32_t pcie_doe_read_config(DOECap *doe_cap, uint32_t addr, int size);
>> +void pcie_doe_write_config(DOECap *doe_cap, uint32_t addr,
>> +                           uint32_t val, int size);
>> +
>> +/* Utility functions for set_rsp in DOEProtocol */
>> +void *pcie_doe_get_req(DOECap *doe_cap);
>> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp);
>> +uint32_t pcie_doe_object_len(void *obj);
>> +
>> +#define DOE_DEBUG
>> +#ifdef DOE_DEBUG
>> +#define DOE_DBG(...) printf(__VA_ARGS__)
>> +#else
>> +#define DOE_DBG(...) {}
>> +#endif
>> +
>> +#define dwsizeof(s) ((sizeof(s) + 4 - 1) / 4)
>> +
>> +#endif /* PCIE_DOE_H */
>> diff --git a/include/hw/pci/pcie_regs.h b/include/hw/pci/pcie_regs.h
>> index 1db86b0..963dc2e 100644
>> --- a/include/hw/pci/pcie_regs.h
>> +++ b/include/hw/pci/pcie_regs.h
>> @@ -179,4 +179,8 @@ typedef enum PCIExpLinkWidth {
>> #define PCI_ACS_VER                     0x1
>> #define PCI_ACS_SIZEOF                  8
>>> +/* DOE Capability Register Fields */
>> +#define PCI_DOE_VER                     0x1
>> +#define PCI_DOE_SIZEOF                  24
>> +
>> #endif /* QEMU_PCIE_REGS_H */
>> diff --git a/include/standard-headers/linux/pci_regs.h b/include/standard-headers/linux/pci_regs.h
>> index e709ae8..4a7b7a4 100644
>> --- a/include/standard-headers/linux/pci_regs.h
>> +++ b/include/standard-headers/linux/pci_regs.h
>> @@ -730,7 +730,8 @@
>> #define PCI_EXT_CAP_ID_DVSEC	0x23	/* Designated Vendor-Specific */
>> #define PCI_EXT_CAP_ID_DLF	0x25	/* Data Link Feature */
>> #define PCI_EXT_CAP_ID_PL_16GT	0x26	/* Physical Layer 16.0 GT/s */
>> -#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_PL_16GT
>> +#define PCI_EXT_CAP_ID_DOE      0x2E    /* Data Object Exchange */
>> +#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_DOE
>> 
>> #define PCI_EXT_CAP_DSN_SIZEOF	12
>> #define PCI_EXT_CAP_MCAST_ENDPOINT_SIZEOF 40
> 
> I haven't reviewed the spec stuff, but I assume Jonathan is familiar with that
> and probably knows it almost by heart already.
> 
>> -- 
>> 1.8.3.1
>> 
>> 

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Jonathan Cameron]

On Tue, 9 Feb 2021 15:35:49 -0500
Chris Browy <cbrowy@avery-design.com> wrote:

Run ./scripts/checkpatch.pl over the patches and fix all the warnings before
posting.  It will save time by clearing out most of the minor formatting issues
and similar that inevitably sneak in during development.

The biggest issue I'm seeing in here is that the abstraction of
multiple DOE capabiltiies accessing same protocols doesn't make sense.

Each DOE ecap region and hence mailbox can have it's own set of
(possibly  overlapping) protocols.

From the ECN:
"It is permitted for a protocol using data object exchanges to require
 that a Function implement a unique instance of DOE for that specific
 protocol, and/or to allow sharing of a DOE instance to only a specific
 set of protocols using data object exchange, and/or to allow a Function
 to implement multiple instances of DOE supporting the specific protocol."

Tightly couple the ECAP and DOE.  If we are in the multiple instances
of DOE supporting a specific protocol case, then register it separately
for each one.  The individual device emulation then needs to deal with
any possible clashes etc.

Various comments inline, but mostly small stuff.

Jonathan


> ---
>  MAINTAINERS                               |   7 +
>  hw/pci/meson.build                        |   1 +
>  hw/pci/pcie.c                             |   2 +-
>  hw/pci/pcie_doe.c                         | 414 ++++++++++++++++++++++++++++++
>  include/hw/pci/pci_ids.h                  |   2 +
>  include/hw/pci/pcie.h                     |   1 +
>  include/hw/pci/pcie_doe.h                 | 166 ++++++++++++
>  include/hw/pci/pcie_regs.h                |   4 +
>  include/standard-headers/linux/pci_regs.h |   3 +-
>  9 files changed, 598 insertions(+), 2 deletions(-)
>  create mode 100644 hw/pci/pcie_doe.c
>  create mode 100644 include/hw/pci/pcie_doe.h
> 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 981dc92..4fb865e 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -1655,6 +1655,13 @@ F: docs/pci*
>  F: docs/specs/*pci*
>  F: default-configs/pci.mak
>  
> +PCIE DOE
> +M: Huai-Cheng Kuo <hchkuo@avery-design.com.tw>
> +M: Chris Browy <cbrowy@avery-design.com>
> +S: Supported
> +F: include/hw/pci/pcie_doe.h
> +F: hw/pci/pcie_doe.c
> +
>  ACPI/SMBIOS
>  M: Michael S. Tsirkin <mst@redhat.com>
>  M: Igor Mammedov <imammedo@redhat.com>
> diff --git a/hw/pci/meson.build b/hw/pci/meson.build
> index 5c4bbac..115e502 100644
> --- a/hw/pci/meson.build
> +++ b/hw/pci/meson.build
> @@ -12,6 +12,7 @@ pci_ss.add(files(
>  # allow plugging PCIe devices into PCI buses, include them even if
>  # CONFIG_PCI_EXPRESS=n.
>  pci_ss.add(files('pcie.c', 'pcie_aer.c'))
> +pci_ss.add(files('pcie_doe.c'))
>  softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: files('pcie_port.c', 'pcie_host.c'))
>  softmmu_ss.add_all(when: 'CONFIG_PCI', if_true: pci_ss)
>  

...

> diff --git a/hw/pci/pcie_doe.c b/hw/pci/pcie_doe.c
> new file mode 100644
> index 0000000..df8e92e
> --- /dev/null
> +++ b/hw/pci/pcie_doe.c
> @@ -0,0 +1,414 @@

Add a copyright header / license etc before v3.

> +#include "qemu/osdep.h"
> +#include "qemu/log.h"
> +#include "qemu/error-report.h"
> +#include "qapi/error.h"
> +#include "qemu/range.h"
> +#include "hw/pci/pci.h"
> +#include "hw/pci/pcie.h"
> +#include "hw/pci/pcie_doe.h"
> +#include "hw/pci/msi.h"
> +#include "hw/pci/msix.h"
> +
> +/*
> + * DOE Default Protocols (Discovery, CMA)
> + */
> +/* Discovery Request Object */
> +struct doe_discovery {
> +    DOEHeader header;
> +    uint8_t index;
> +    uint8_t reserved[3];
> +} QEMU_PACKED;
> +
> +/* Discovery Response Object */
> +struct doe_discovery_rsp {
> +    DOEHeader header;
> +    uint16_t vendor_id;
> +    uint8_t doe_type;
> +    uint8_t next_index;
> +} QEMU_PACKED;
> +
> +/* Callback for Discovery */
> +static bool pcie_doe_discovery_rsp(DOECap *doe_cap)
> +{
> +    PCIEDOE *doe = doe_cap->doe;
> +    struct doe_discovery *req = pcie_doe_get_req(doe_cap);
> +    uint8_t index = req->index;
> +    DOEProtocol *prot = NULL;
> +
> +    /* Request length mismatch, discard */
> +    if (req->header.length < dwsizeof(struct doe_discovery)) {
> +        return DOE_DISCARD;

	return false;  Or better yet a meaningful error code to make debugging
easier.

> +    }
> +
> +    /* Point to the requested protocol */
> +    if (index < doe->protocol_num) {
> +        prot = &doe->protocols[index];
> +    }
> +
> +    struct doe_discovery_rsp rsp = {
> +        .header = {
> +            .vendor_id = PCI_VENDOR_ID_PCI_SIG,
> +            .doe_type = PCI_SIG_DOE_DISCOVERY,
> +            .reserved = 0x0,

Nice thing about c99 based structure assignment is that unspecified
elements are set to 0 automatically.  So you can drop this particular
element safely and end up with slightly cleaner code.

> +            .length = dwsizeof(struct doe_discovery_rsp),
> +        },
> +        .vendor_id = (prot) ? prot->vendor_id : 0xFFFF,
> +        .doe_type = (prot) ? prot->doe_type : 0xFF,
> +        .next_index = (index + 1) < doe->protocol_num ?
> +                      (index + 1) : 0,
> +    };
> +
> +    pcie_doe_set_rsp(doe_cap, &rsp);
> +
> +    return DOE_SUCCESS;
> +}
> +
> +/* Callback for CMA */
> +static bool pcie_doe_cma_rsp(DOECap *doe_cap)

I've not checked CMA, but I'd expect this to be a separate
patch as not part of core DOE support (or same ECN for that
matter).

> +{
> +    doe_cap->status.error = 1;
> +
> +    memset(doe_cap->read_mbox, 0,
> +           PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +
> +    doe_cap->write_mbox_len = 0;
> +
> +    return DOE_DISCARD;
> +}
> +
> +/*
> + * DOE Utilities
> + */
> +static void pcie_doe_reset_mbox(DOECap *st)
> +{
> +    st->read_mbox_idx = 0;
> +
> +    st->read_mbox_len = 0;
> +    st->write_mbox_len = 0;
> +
> +    memset(st->read_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +    memset(st->write_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +}
> +
> +/*
> + * Initialize the list and protocol for a device.
> + * This function won't add the DOE capabitity to your PCIe device.
> + */
> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe)
> +{
> +    doe->pdev = dev;
> +    doe->head = NULL;
> +    doe->protocol_num = 0;

No need to set things to zero as I assume you allocate it zero filled.
At least no point for things where zero is the obvious default.

> +
> +    /* Register two default protocol */
> +    //TODO : LINK LIST

Agreed :)

> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
> +            PCI_SIG_DOE_DISCOVERY, pcie_doe_discovery_rsp);
> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
> +            PCI_SIG_DOE_CMA, pcie_doe_cma_rsp);
> +}
> +
> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec) {

Bracket on this line.

> +    DOECap *new_cap, **ptr;
> +    PCIDevice *dev = doe->pdev;
> +
> +    pcie_add_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_VER, offset,
> +                        PCI_DOE_SIZEOF);
> +
> +    ptr = &doe->head;
> +    /* Insert the new DOE at the end of linked list */

As below, not sure this abstraction makes sense.

> +    while (*ptr) {
> +        if (range_covers_byte((*ptr)->offset, PCI_DOE_SIZEOF, offset) ||
> +            (*ptr)->cap.vec == vec) {
> +            return -EINVAL;
> +        }
> +
> +        ptr = &(*ptr)->next;
> +    }
> +    new_cap = g_malloc0(sizeof(DOECap));
> +    *ptr = new_cap;
> +
> +    new_cap->doe = doe;
> +    new_cap->offset = offset;
> +    new_cap->next = NULL;
> +
> +    /* Configure MSI/MSI-X */
> +    if (intr && (msi_present(dev) || msix_present(dev))) {
> +        new_cap->cap.intr = intr;
> +        new_cap->cap.vec = vec;
> +    }
> +
> +    /* Set up W/R Mailbox buffer */
> +    new_cap->write_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +    new_cap->read_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
> +
> +    pcie_doe_reset_mbox(new_cap);
> +
> +    return 0;
> +}
> +
> +void pcie_doe_uninit(PCIEDOE *doe) {
> +    PCIDevice *dev = doe->pdev;
> +    DOECap *cap;
> +
> +    pci_del_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_SIZEOF);
> +
> +    cap = doe->head;
> +    while (cap) {
> +        doe->head = doe->head->next;
> +
> +        g_free(cap->read_mbox);
> +        g_free(cap->write_mbox);
> +        cap->read_mbox = NULL;
> +        cap->write_mbox = NULL;

I'd avoid setting things to NULL that you are throwing away.  It
implies that they will be reused or that it matters in somewhat which
isn't the case here.

> +        g_free(cap);
> +        cap = doe->head;
> +    }
> +}
> +
> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
> +        uint8_t doe_type, bool (*set_rsp)(DOECap *))
> +{
> +    /* Protocol num should not exceed 256 */
> +    assert(doe->protocol_num < PCI_DOE_PROTOCOL_MAX);
> +
> +    doe->protocols[doe->protocol_num].vendor_id = vendor_id;
> +    doe->protocols[doe->protocol_num].doe_type = doe_type;
> +    doe->protocols[doe->protocol_num].set_rsp = set_rsp;
> +
> +    doe->protocol_num++;
> +}
> +
> +uint32_t pcie_doe_build_protocol(DOEProtocol *p)
> +{
> +    return DATA_OBJ_BUILD_HEADER1(p->vendor_id, p->doe_type);
> +}
> +
> +/* Return the pointer of DOE request in write mailbox buffer */
> +void *pcie_doe_get_req(DOECap *doe_cap)
> +{
> +    return doe_cap->write_mbox;
> +}
> +
> +/* Copy the response to read mailbox buffer */
> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp)
> +{
> +    uint32_t len = pcie_doe_object_len(rsp);
> +
> +    memcpy(doe_cap->read_mbox + doe_cap->read_mbox_len,
> +           rsp, len * sizeof(uint32_t));
> +    doe_cap->read_mbox_len += len;
> +}
> +
> +/* Get Data Object length */
> +uint32_t pcie_doe_object_len(void *obj)
> +{
> +    return (obj)? ((DOEHeader *)obj)->length : 0;
> +}
> +
> +static void pcie_doe_write_mbox(DOECap *doe_cap, uint32_t val)
> +{
> +    memcpy(doe_cap->write_mbox + doe_cap->write_mbox_len, &val, sizeof(uint32_t));
> +
> +    if (doe_cap->write_mbox_len == 1) {
> +        DOE_DBG("  Capture DOE request DO length = %d\n", val & 0x0003ffff);
> +    }
> +
> +    doe_cap->write_mbox_len++;

We probably need to check that the mailbox write was full dword as the spec
requires.  No idea what we do if it isn't as spec doesn't seem to say...
I've queried one of our PCI experts.


> +}
> +
> +static void pcie_doe_irq_assert(DOECap *doe_cap)
> +{
> +    PCIDevice *dev = doe_cap->doe->pdev;
> +
> +    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {
> +        /* Interrupt notify */
> +        if (msix_enabled(dev)) {
> +            msix_notify(dev, doe_cap->cap.vec);
> +        } else if (msi_enabled(dev)) {
> +            msi_notify(dev, doe_cap->cap.vec);
> +        }
> +        /* Not support legacy IRQ */
> +    }
> +}
> +
> +static void pcie_doe_set_ready(DOECap *doe_cap, bool rdy)
> +{
> +    doe_cap->status.ready = rdy;
> +
> +    if (rdy) {
> +        pcie_doe_irq_assert(doe_cap);
> +    }
> +}
> +
> +static void pcie_doe_set_error(DOECap *doe_cap, bool err)
> +{
> +    doe_cap->status.error = err;
> +
> +    if (err) {
> +        pcie_doe_irq_assert(doe_cap);
> +    }
> +}
> +
> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr)
> +{
> +    DOECap *ptr = doe->head;
> +
> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
> +    while (ptr && 
> +           !range_covers_byte(ptr->offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
> +        ptr = ptr->next;
> +    }
> +    
> +    return ptr;
> +}
> +
> +uint32_t pcie_doe_read_config(DOECap *doe_cap,
> +                              uint32_t addr, int size)
> +{
> +    uint32_t ret = 0, shift, mask = 0xFFFFFFFF;
> +    uint16_t doe_offset = doe_cap->offset;
> +
> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {

I'd flip this around to reduce indentation + no need to be careful with alignment etc
if we aren't returning anything.

       if (!range_cover_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)
		return 0;


> +        addr -= doe_offset;
> +
> +        if (range_covers_byte(PCI_EXP_DOE_CAP, sizeof(uint32_t), addr)) {
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, INTR_SUPP,
> +                             doe_cap->cap.intr);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM,
> +                             doe_cap->cap.vec);
> +        } else if (range_covers_byte(PCI_EXP_DOE_CTRL, sizeof(uint32_t), addr)) {
> +            /* Must return ABORT=0 and GO=0 */
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_CONTROL, DOE_INTR_EN,
> +                             doe_cap->ctrl.intr);
> +            DOE_DBG("  CONTROL REG=%x\n", ret);
> +        } else if (range_covers_byte(PCI_EXP_DOE_STATUS, sizeof(uint32_t), addr)) {
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_BUSY,
> +                             doe_cap->status.busy);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_INTR_STATUS,
> +                             doe_cap->status.intr);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_ERROR,
> +                             doe_cap->status.error);
> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DATA_OBJ_RDY,
> +                             doe_cap->status.ready);
> +        } else if (range_covers_byte(PCI_EXP_DOE_RD_DATA_MBOX, sizeof(uint32_t), addr)) {
> +            /* Check that READY is set */

I'd clean out any comment that is obvious from the code.
Comments get out of sync over time so there is a maintenance burden in having them such
that we only bother if they add information.

> +            if (!doe_cap->status.ready) {
> +                /* Return 0 if not ready */
> +                DOE_DBG("  RD MBOX RETURN=%x\n", ret);
> +            } else {
> +                /* Deposit next DO DW into read mbox */

This comment is missleading.  It might not be the 'next' one. If you read
the register twice for instance.  Much better not to have the comment
at all on basis code is fairly obvious anyway!

As mentioned below, a read of this when nothing there is an underflow and spec
suggests that is an error condition.

> +                DOE_DBG("  RD MBOX, DATA OBJECT READY,"
> +                        " CURRENT DO DW OFFSET=%x\n",
> +                        doe_cap->read_mbox_idx);
> +
> +                ret = doe_cap->read_mbox[doe_cap->read_mbox_idx];
> +
> +                DOE_DBG("  RD MBOX DW=%x\n", ret);
> +                DOE_DBG("  RD MBOX DW OFFSET=%d, RD MBOX LENGTH=%d\n",
> +                        doe_cap->read_mbox_idx, doe_cap->read_mbox_len);
> +            }
> +        } else if (range_covers_byte(PCI_EXP_DOE_WR_DATA_MBOX, sizeof(uint32_t), addr)) {
> +            DOE_DBG("  WR MBOX, local_val =%x\n", ret);
> +        }
> +    }
> +
> +    /* Alignment */
> +    shift = addr % sizeof(uint32_t);
> +    if (shift) {
> +        ret >>= shift * 8;
> +    }
> +    mask >>= (sizeof(uint32_t) - size) * 8;
> +
> +    return ret & mask;
> +}
> +
> +void pcie_doe_write_config(DOECap *doe_cap,
> +                           uint32_t addr, uint32_t val, int size)
> +{
> +    PCIEDOE *doe = doe_cap->doe;
> +    uint16_t doe_offset = doe_cap->offset;
> +    int p;
> +    bool discard;
> +    uint32_t shift;
> +
> +    DOE_DBG("  addr=%x, val=%x, size=%x\n", addr, val, size);
> +
> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {

As below, invert this condition and return early as it will simplify below and reduce
indentation.

     if (!range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
          return;
    }

> +        /* Alignment */
> +        shift = addr % sizeof(uint32_t);
> +        addr -= (doe_offset - shift);
> +        val <<= shift * 8;
> +
> +        switch (addr) {
> +        case PCI_EXP_DOE_CTRL:
> +            DOE_DBG("  CONTROL=%x\n", val);
> +            if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_ABORT)) {
> +                /* If ABORT, clear status reg DOE Error and DOE Ready */
> +                DOE_DBG("  Setting ABORT\n");
> +                pcie_doe_set_ready(doe_cap, 0);
> +                pcie_doe_set_error(doe_cap, 0);
> +                pcie_doe_reset_mbox(doe_cap);
> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_GO)) {
> +                DOE_DBG("  CONTROL GO=%x\n", val);

This case is complex enough I'd factor it out as it's own function.

> +                /*
> +                 * Check protocol the incoming request in write_mbox and
> +                 * memcpy the corresponding response to read_mbox.
> +                 *
> +                 * "discard" should be set up if the response callback
> +                 * function could not deal with request (e.g. length
> +                 * mismatch) or the protocol of request was not found.
> +                 */
> +                discard = DOE_DISCARD;
> +                for (p = 0; p < doe->protocol_num; p++) {
> +                    if (doe_cap->write_mbox[0] ==
> +                        pcie_doe_build_protocol(&doe->protocols[p])) {
> +                        /* Found */
> +                        DOE_DBG("  DOE PROTOCOL=%x\n", doe_cap->write_mbox[0]);
> +                        /*
> +                         * Spec:
> +                         * If the number of DW transferred does not match the
> +                         * indicated Length for a data object, then the
> +                         * data object must be silently discarded.
> +                         */
> +                        if (doe_cap->write_mbox_len ==
> +                            pcie_doe_object_len(pcie_doe_get_req(doe_cap)))
> +                            discard = doe->protocols[p].set_rsp(doe_cap);
> +                        break;
> +                    }
> +                }
> +
> +                /* Set DOE Ready */
> +                if (!discard) {
> +                    pcie_doe_set_ready(doe_cap, 1);
> +                } else {
> +                    pcie_doe_reset_mbox(doe_cap);
> +                }
> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_INTR_EN)) {

Spec reference needed to say why you can't write this at same time as GO.
It may be odd thing to do, but I can't see anything saying you couldn't do this,
for example on your very first command.

> +                doe_cap->ctrl.intr = 1;
> +            }
> +            break;
> +        case PCI_EXP_DOE_RD_DATA_MBOX:
> +            /* Read MBOX */
> +            DOE_DBG("  INCR RD MBOX DO DW OFFSET=%d\n", doe_cap->read_mbox_idx);
> +            doe_cap->read_mbox_idx += 1;
> +            /* Last DW */
> +            if (doe_cap->read_mbox_idx >= doe_cap->read_mbox_len) {
> +                pcie_doe_reset_mbox(doe_cap);
> +                pcie_doe_set_ready(doe_cap, 0);
> +            }

A write of this when there is nothing there is an underflow. As is a read of this
after the last write.  I would guess both should be error conditions.

> +            break;
> +        case PCI_EXP_DOE_WR_DATA_MBOX:
> +            /* Write MBOX */
> +            DOE_DBG("  WR MBOX=%x, DW OFFSET = %d\n", val, doe_cap->write_mbox_len);
> +            pcie_doe_write_mbox(doe_cap, val);
> +            break;
> +        case PCI_EXP_DOE_STATUS:
> +        case PCI_EXP_DOE_CAP:
> +        default:
> +            break;
> +        }
> +    }
> +}
> diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
> index 76bf3ed..636b2e8 100644
> --- a/include/hw/pci/pci_ids.h
> +++ b/include/hw/pci/pci_ids.h
> @@ -157,6 +157,8 @@
>  
>  /* Vendors and devices.  Sort key: vendor first, device next. */
>  
> +#define PCI_VENDOR_ID_PCI_SIG            0x0001
> +
>  #define PCI_VENDOR_ID_LSI_LOGIC          0x1000
>  #define PCI_DEVICE_ID_LSI_53C810         0x0001
>  #define PCI_DEVICE_ID_LSI_53C895A        0x0012
> diff --git a/include/hw/pci/pcie.h b/include/hw/pci/pcie.h
> index 14c58eb..47d6f66 100644
> --- a/include/hw/pci/pcie.h
> +++ b/include/hw/pci/pcie.h
> @@ -25,6 +25,7 @@
>  #include "hw/pci/pcie_regs.h"
>  #include "hw/pci/pcie_aer.h"
>  #include "hw/hotplug.h"
> +#include "hw/pci/pcie_doe.h"
>  
>  typedef enum {
>      /* for attention and power indicator */
> diff --git a/include/hw/pci/pcie_doe.h b/include/hw/pci/pcie_doe.h
> new file mode 100644
> index 0000000..f497075
> --- /dev/null
> +++ b/include/hw/pci/pcie_doe.h
> @@ -0,0 +1,166 @@
> +#ifndef PCIE_DOE_H
> +#define PCIE_DOE_H
> +
> +#include "qemu/range.h"
> +#include "qemu/typedefs.h"
> +#include "hw/register.h"
> +
> +/* 
> + * PCI DOE register defines 7.9.xx 
> + */
> +/* DOE Capabilities Register */
> +#define PCI_EXP_DOE_CAP             0x04
> +REG32(PCI_DOE_CAP_REG, 0)
> +    FIELD(PCI_DOE_CAP_REG, INTR_SUPP, 0, 1)
> +    FIELD(PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM, 1, 11)
> +
> +/* DOE Control Register */
> +#define PCI_EXP_DOE_CTRL            0x08
> +REG32(PCI_DOE_CAP_CONTROL, 0)
> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_ABORT, 0, 1)
> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_INTR_EN, 1, 1)
> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_GO, 31, 1)
> +
> +/* DOE Status Register  */
> +#define PCI_EXP_DOE_STATUS          0x0c
> +REG32(PCI_DOE_CAP_STATUS, 0)
> +    FIELD(PCI_DOE_CAP_STATUS, DOE_BUSY, 0, 1)
> +    FIELD(PCI_DOE_CAP_STATUS, DOE_INTR_STATUS, 1, 1)
> +    FIELD(PCI_DOE_CAP_STATUS, DOE_ERROR, 2, 1)
> +    FIELD(PCI_DOE_CAP_STATUS, DATA_OBJ_RDY, 31, 1)
> +
> +/* DOE Write Data Mailbox Register  */
> +#define PCI_EXP_DOE_WR_DATA_MBOX    0x10
> +
> +/* DOE Read Data Mailbox Register  */
> +#define PCI_EXP_DOE_RD_DATA_MBOX    0x14
> +
> +/* 
> + * PCI DOE register defines 7.9.xx 
> + */
> +/* Table 7-x2 */
> +#define PCI_SIG_DOE_DISCOVERY       0x00
> +#define PCI_SIG_DOE_CMA             0x01
> +
> +#define DATA_OBJ_BUILD_HEADER1(v, p)  ((p << 16) | v)
> +
> +/* Figure 6-x1 */
> +#define DATA_OBJECT_HEADER1_OFFSET  0
> +#define DATA_OBJECT_HEADER2_OFFSET  1
> +#define DATA_OBJECT_CONTENT_OFFSET  2
> +
> +#define PCI_DOE_MAX_DW_SIZE (1 << 18)
> +#define PCI_DOE_PROTOCOL_MAX 256
> +
> +/*
> + * Self-defined Marco
> + */
> +/* Request/Response State */
> +#define DOE_SUCCESS 0
> +#define DOE_DISCARD 1
Drop these. As mentioned in previous review.  These are just
obvious bools.

> +
> +/* An invalid vector number for MSI/MSI-x */
> +#define DOE_NO_INTR (~0)
> +
> +/*
> + * DOE Protocol - Data Object
> + */
> +typedef struct DOEHeader DOEHeader;
> +typedef struct DOEProtocol DOEProtocol;
> +typedef struct PCIEDOE PCIEDOE;
> +typedef struct DOECap DOECap;
> +
> +struct DOEHeader {
> +    uint16_t vendor_id;
> +    uint8_t doe_type;
> +    uint8_t reserved;
> +    struct {
> +        uint32_t length:18;
> +        uint32_t reserved2:14;

As before, bitfields are not a good idea in this sort of code in general due to lack
of standard handling in the C spec.

> +    };
> +} QEMU_PACKED;
> +
> +/* Protocol infos and rsp function callback */
> +struct DOEProtocol {
> +    uint16_t vendor_id;
> +    uint8_t doe_type;
> +
> +    bool (*set_rsp)(DOECap *);
> +};
> +
> +struct PCIEDOE {
> +    PCIDevice *pdev;

Having the PCIDevice in here only allows you to drop a parameter in read and write
calls. I'd not bother given the nest of pointers you have as a result.
Just pass both the PCIDOE and the PCIDevice into those two functions.

> +    DOECap *head;

I can sort of get what you are doing with this list of DOEs but to mind it's the wrong
abstraction as it doesn't fit how I think these will be used.

It's not a case that there will be N DOE mailboxes each of which supports
all the same protocols.  I suspect quite the opposite.

Each DOE may support multiple protocols but the most obvious reason you'd
do multiple DOEs is because they support different protocols.

If we want to support the same protocol on mulitple DOE instances then we'd
register it for each of them.

Thus I'd drop the list handling and instead  have for example

struct cxl_type3_dev {

...

    PCIDOE doe_ima;
    PCIDOE doe_cdat;
};

In the read and write calls then just call pcie_doe_xxxx_config() twice.
You do have to handle the miss vs hit stuff though (similar problem that
lead to ugly code in my version).


> +
> +    /* Protocols and its callback response */
> +    DOEProtocol protocols[PCI_DOE_PROTOCOL_MAX];
> +    uint32_t protocol_num;
> +};
> +
> +struct DOECap {
> +    PCIEDOE *doe;

Following suggestion to get rid of the list, you should also then combine
PCIEDOE and the DOECap as they match 1 to 1.

> +
> +    /* Capability offset */
> +    uint16_t offset;
> +
> +    /* Next DOE capability */
> +    DOECap *next;
> +
> +    /* Capability */
> +    struct {
> +        bool intr;
> +        uint16_t vec;
> +    } cap;
> +
> +    /* Control */
> +    struct {
> +        bool abort;
> +        bool intr;
> +        bool go;
> +    } ctrl;
> +
> +    /* Status */
> +    struct {
> +        bool busy;
> +        bool intr;
> +        bool error;
> +        bool ready;
> +    } status;
> +
> +    /* Mailbox buffer for device */
> +    uint32_t *write_mbox;
> +    uint32_t *read_mbox;
> +
> +    /* Mailbox position indicator */
> +    uint32_t read_mbox_idx;
> +    uint32_t read_mbox_len;
> +    uint32_t write_mbox_len;
> +};
> +
> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe);
> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec);
> +void pcie_doe_uninit(PCIEDOE *doe);
> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
> +                                uint8_t doe_type,
> +                                bool (*set_rsp)(DOECap *));
> +uint32_t pcie_doe_build_protocol(DOEProtocol *p);
> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr);
> +uint32_t pcie_doe_read_config(DOECap *doe_cap, uint32_t addr, int size);
> +void pcie_doe_write_config(DOECap *doe_cap, uint32_t addr,
> +                           uint32_t val, int size);
> +
> +/* Utility functions for set_rsp in DOEProtocol */
> +void *pcie_doe_get_req(DOECap *doe_cap);
> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp);
> +uint32_t pcie_doe_object_len(void *obj);
> +
> +#define DOE_DEBUG
> +#ifdef DOE_DEBUG
> +#define DOE_DBG(...) printf(__VA_ARGS__)
> +#else
> +#define DOE_DBG(...) {}
> +#endif

Tidy this stuff up (i.e. get rid of it) for next version.  It's fine when
you are doing early debug, but we don't want it in the version for review.

> +
> +#define dwsizeof(s) ((sizeof(s) + 4 - 1) / 4)

Not used in this patch.  Move it down towards code that uses it or at very 
least only introduce it when used.

> +
> +#endif /* PCIE_DOE_H */
> diff --git a/include/hw/pci/pcie_regs.h b/include/hw/pci/pcie_regs.h
> index 1db86b0..963dc2e 100644
> --- a/include/hw/pci/pcie_regs.h
> +++ b/include/hw/pci/pcie_regs.h
> @@ -179,4 +179,8 @@ typedef enum PCIExpLinkWidth {
>  #define PCI_ACS_VER                     0x1
>  #define PCI_ACS_SIZEOF                  8
>  
> +/* DOE Capability Register Fields */
> +#define PCI_DOE_VER                     0x1
> +#define PCI_DOE_SIZEOF                  24
> +
>  #endif /* QEMU_PCIE_REGS_H */
> diff --git a/include/standard-headers/linux/pci_regs.h b/include/standard-headers/linux/pci_regs.h
> index e709ae8..4a7b7a4 100644
> --- a/include/standard-headers/linux/pci_regs.h
> +++ b/include/standard-headers/linux/pci_regs.h

Pull this change out as a separate patch.  This header gets fetched via a script
so we will only find this here if we update the source file in the kernel.

That should happen soon anyway as we add the support to read this.


> @@ -730,7 +730,8 @@
>  #define PCI_EXT_CAP_ID_DVSEC	0x23	/* Designated Vendor-Specific */
>  #define PCI_EXT_CAP_ID_DLF	0x25	/* Data Link Feature */
>  #define PCI_EXT_CAP_ID_PL_16GT	0x26	/* Physical Layer 16.0 GT/s */
> -#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_PL_16GT
> +#define PCI_EXT_CAP_ID_DOE      0x2E    /* Data Object Exchange */
> +#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_DOE
>  
>  #define PCI_EXT_CAP_DSN_SIZEOF	12
>  #define PCI_EXT_CAP_MCAST_ENDPOINT_SIZEOF 40

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Chris Browy]

> On Feb 12, 2021, at 11:24 AM, Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:
> 
> On Tue, 9 Feb 2021 15:35:49 -0500
> Chris Browy <cbrowy@avery-design.com> wrote:
> 
> Run ./scripts/checkpatch.pl over the patches and fix all the warnings before
> posting.  It will save time by clearing out most of the minor formatting issues
> and similar that inevitably sneak in during development.
> 
Excellent suggestion.  We’re still newbies!

> The biggest issue I'm seeing in here is that the abstraction of
> multiple DOE capabiltiies accessing same protocols doesn't make sense.
> 
> Each DOE ecap region and hence mailbox can have it's own set of
> (possibly  overlapping) protocols.
> 
> From the ECN:
> "It is permitted for a protocol using data object exchanges to require
> that a Function implement a unique instance of DOE for that specific
> protocol, and/or to allow sharing of a DOE instance to only a specific
> set of protocols using data object exchange, and/or to allow a Function
> to implement multiple instances of DOE supporting the specific protocol."
> 
> Tightly couple the ECAP and DOE.  If we are in the multiple instances
> of DOE supporting a specific protocol case, then register it separately
> for each one.  The individual device emulation then needs to deal with
> any possible clashes etc.

Not sure how configurable we want to make the device.  It is a simple type 3
device after all. 

The DOE spec does leave it pretty arbitrary regarding N DOE instances (DOE 
Extended Cap entry points) for M protocols, including where N>1 and M=1.  
Currently we implement N=2 DOE caps (instances), one for CDAT, one for 
Compliance Mode.

Maybe a more complex MLD device might have one or more DOE instances 
for the CDAT protocol alone to define each HDM but currently we only have 
one pmem (SLD) so we can’t really do much more than what’s supported.

Open to further suggestion though.  Based on answer to above we’ll follow 
the suggestion lower in the code review regarding 


> 
> Various comments inline, but mostly small stuff.
> 
> Jonathan
> 
> 
>> ---
>> MAINTAINERS                               |   7 +
>> hw/pci/meson.build                        |   1 +
>> hw/pci/pcie.c                             |   2 +-
>> hw/pci/pcie_doe.c                         | 414 ++++++++++++++++++++++++++++++
>> include/hw/pci/pci_ids.h                  |   2 +
>> include/hw/pci/pcie.h                     |   1 +
>> include/hw/pci/pcie_doe.h                 | 166 ++++++++++++
>> include/hw/pci/pcie_regs.h                |   4 +
>> include/standard-headers/linux/pci_regs.h |   3 +-
>> 9 files changed, 598 insertions(+), 2 deletions(-)
>> create mode 100644 hw/pci/pcie_doe.c
>> create mode 100644 include/hw/pci/pcie_doe.h
>> 
>> diff --git a/MAINTAINERS b/MAINTAINERS
>> index 981dc92..4fb865e 100644
>> --- a/MAINTAINERS
>> +++ b/MAINTAINERS
>> @@ -1655,6 +1655,13 @@ F: docs/pci*
>> F: docs/specs/*pci*
>> F: default-configs/pci.mak
>> 
>> +PCIE DOE
>> +M: Huai-Cheng Kuo <hchkuo@avery-design.com.tw>
>> +M: Chris Browy <cbrowy@avery-design.com>
>> +S: Supported
>> +F: include/hw/pci/pcie_doe.h
>> +F: hw/pci/pcie_doe.c
>> +
>> ACPI/SMBIOS
>> M: Michael S. Tsirkin <mst@redhat.com>
>> M: Igor Mammedov <imammedo@redhat.com>
>> diff --git a/hw/pci/meson.build b/hw/pci/meson.build
>> index 5c4bbac..115e502 100644
>> --- a/hw/pci/meson.build
>> +++ b/hw/pci/meson.build
>> @@ -12,6 +12,7 @@ pci_ss.add(files(
>> # allow plugging PCIe devices into PCI buses, include them even if
>> # CONFIG_PCI_EXPRESS=n.
>> pci_ss.add(files('pcie.c', 'pcie_aer.c'))
>> +pci_ss.add(files('pcie_doe.c'))
>> softmmu_ss.add(when: 'CONFIG_PCI_EXPRESS', if_true: files('pcie_port.c', 'pcie_host.c'))
>> softmmu_ss.add_all(when: 'CONFIG_PCI', if_true: pci_ss)
>> 
> 
> ...
> 
>> diff --git a/hw/pci/pcie_doe.c b/hw/pci/pcie_doe.c
>> new file mode 100644
>> index 0000000..df8e92e
>> --- /dev/null
>> +++ b/hw/pci/pcie_doe.c
>> @@ -0,0 +1,414 @@
> 
> Add a copyright header / license etc before v3.
> 
>> +#include "qemu/osdep.h"
>> +#include "qemu/log.h"
>> +#include "qemu/error-report.h"
>> +#include "qapi/error.h"
>> +#include "qemu/range.h"
>> +#include "hw/pci/pci.h"
>> +#include "hw/pci/pcie.h"
>> +#include "hw/pci/pcie_doe.h"
>> +#include "hw/pci/msi.h"
>> +#include "hw/pci/msix.h"
>> +
>> +/*
>> + * DOE Default Protocols (Discovery, CMA)
>> + */
>> +/* Discovery Request Object */
>> +struct doe_discovery {
>> +    DOEHeader header;
>> +    uint8_t index;
>> +    uint8_t reserved[3];
>> +} QEMU_PACKED;
>> +
>> +/* Discovery Response Object */
>> +struct doe_discovery_rsp {
>> +    DOEHeader header;
>> +    uint16_t vendor_id;
>> +    uint8_t doe_type;
>> +    uint8_t next_index;
>> +} QEMU_PACKED;
>> +
>> +/* Callback for Discovery */
>> +static bool pcie_doe_discovery_rsp(DOECap *doe_cap)
>> +{
>> +    PCIEDOE *doe = doe_cap->doe;
>> +    struct doe_discovery *req = pcie_doe_get_req(doe_cap);
>> +    uint8_t index = req->index;
>> +    DOEProtocol *prot = NULL;
>> +
>> +    /* Request length mismatch, discard */
>> +    if (req->header.length < dwsizeof(struct doe_discovery)) {
>> +        return DOE_DISCARD;
> 
> 	return false;  Or better yet a meaningful error code to make debugging
> easier.

OK

> 
>> +    }
>> +
>> +    /* Point to the requested protocol */
>> +    if (index < doe->protocol_num) {
>> +        prot = &doe->protocols[index];
>> +    }
>> +
>> +    struct doe_discovery_rsp rsp = {
>> +        .header = {
>> +            .vendor_id = PCI_VENDOR_ID_PCI_SIG,
>> +            .doe_type = PCI_SIG_DOE_DISCOVERY,
>> +            .reserved = 0x0,
> 
> Nice thing about c99 based structure assignment is that unspecified
> elements are set to 0 automatically.  So you can drop this particular
> element safely and end up with slightly cleaner code.

OK

> 
>> +            .length = dwsizeof(struct doe_discovery_rsp),
>> +        },
>> +        .vendor_id = (prot) ? prot->vendor_id : 0xFFFF,
>> +        .doe_type = (prot) ? prot->doe_type : 0xFF,
>> +        .next_index = (index + 1) < doe->protocol_num ?
>> +                      (index + 1) : 0,
>> +    };
>> +
>> +    pcie_doe_set_rsp(doe_cap, &rsp);
>> +
>> +    return DOE_SUCCESS;
>> +}
>> +
>> +/* Callback for CMA */
>> +static bool pcie_doe_cma_rsp(DOECap *doe_cap)
> 
> I've not checked CMA, but I'd expect this to be a separate
> patch as not part of core DOE support (or same ECN for that
> matter).

We’ll back out CMA.  Currently it’s an incomplete solution.  We support it in 
discovery process but the response is that of being disabled.  So not the best
way to handle it.

> 
>> +{
>> +    doe_cap->status.error = 1;
>> +
>> +    memset(doe_cap->read_mbox, 0,
>> +           PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +
>> +    doe_cap->write_mbox_len = 0;
>> +
>> +    return DOE_DISCARD;
>> +}
>> +
>> +/*
>> + * DOE Utilities
>> + */
>> +static void pcie_doe_reset_mbox(DOECap *st)
>> +{
>> +    st->read_mbox_idx = 0;
>> +
>> +    st->read_mbox_len = 0;
>> +    st->write_mbox_len = 0;
>> +
>> +    memset(st->read_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +    memset(st->write_mbox, 0, PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +}
>> +
>> +/*
>> + * Initialize the list and protocol for a device.
>> + * This function won't add the DOE capabitity to your PCIe device.
>> + */
>> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe)
>> +{
>> +    doe->pdev = dev;
>> +    doe->head = NULL;
>> +    doe->protocol_num = 0;
> 
> No need to set things to zero as I assume you allocate it zero filled.
> At least no point for things where zero is the obvious default.
> 
>> +
>> +    /* Register two default protocol */
>> +    //TODO : LINK LIST
> 
> Agreed :)
> 
>> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
>> +            PCI_SIG_DOE_DISCOVERY, pcie_doe_discovery_rsp);
>> +    pcie_doe_register_protocol(doe, PCI_VENDOR_ID_PCI_SIG,
>> +            PCI_SIG_DOE_CMA, pcie_doe_cma_rsp);
>> +}
>> +
>> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec) {
> 
> Bracket on this line.
> 
>> +    DOECap *new_cap, **ptr;
>> +    PCIDevice *dev = doe->pdev;
>> +
>> +    pcie_add_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_VER, offset,
>> +                        PCI_DOE_SIZEOF);
>> +
>> +    ptr = &doe->head;
>> +    /* Insert the new DOE at the end of linked list */
> 
> As below, not sure this abstraction makes sense.
> 
>> +    while (*ptr) {
>> +        if (range_covers_byte((*ptr)->offset, PCI_DOE_SIZEOF, offset) ||
>> +            (*ptr)->cap.vec == vec) {
>> +            return -EINVAL;
>> +        }
>> +
>> +        ptr = &(*ptr)->next;
>> +    }
>> +    new_cap = g_malloc0(sizeof(DOECap));
>> +    *ptr = new_cap;
>> +
>> +    new_cap->doe = doe;
>> +    new_cap->offset = offset;
>> +    new_cap->next = NULL;
>> +
>> +    /* Configure MSI/MSI-X */
>> +    if (intr && (msi_present(dev) || msix_present(dev))) {
>> +        new_cap->cap.intr = intr;
>> +        new_cap->cap.vec = vec;
>> +    }
>> +
>> +    /* Set up W/R Mailbox buffer */
>> +    new_cap->write_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +    new_cap->read_mbox = g_malloc0(PCI_DOE_MAX_DW_SIZE * sizeof(uint32_t));
>> +
>> +    pcie_doe_reset_mbox(new_cap);
>> +
>> +    return 0;
>> +}
>> +
>> +void pcie_doe_uninit(PCIEDOE *doe) {
>> +    PCIDevice *dev = doe->pdev;
>> +    DOECap *cap;
>> +
>> +    pci_del_capability(dev, PCI_EXT_CAP_ID_DOE, PCI_DOE_SIZEOF);
>> +
>> +    cap = doe->head;
>> +    while (cap) {
>> +        doe->head = doe->head->next;
>> +
>> +        g_free(cap->read_mbox);
>> +        g_free(cap->write_mbox);
>> +        cap->read_mbox = NULL;
>> +        cap->write_mbox = NULL;
> 
> I'd avoid setting things to NULL that you are throwing away.  It
> implies that they will be reused or that it matters in somewhat which
> isn't the case here.

OK

> 
>> +        g_free(cap);
>> +        cap = doe->head;
>> +    }
>> +}
>> +
>> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
>> +        uint8_t doe_type, bool (*set_rsp)(DOECap *))
>> +{
>> +    /* Protocol num should not exceed 256 */
>> +    assert(doe->protocol_num < PCI_DOE_PROTOCOL_MAX);
>> +
>> +    doe->protocols[doe->protocol_num].vendor_id = vendor_id;
>> +    doe->protocols[doe->protocol_num].doe_type = doe_type;
>> +    doe->protocols[doe->protocol_num].set_rsp = set_rsp;
>> +
>> +    doe->protocol_num++;
>> +}
>> +
>> +uint32_t pcie_doe_build_protocol(DOEProtocol *p)
>> +{
>> +    return DATA_OBJ_BUILD_HEADER1(p->vendor_id, p->doe_type);
>> +}
>> +
>> +/* Return the pointer of DOE request in write mailbox buffer */
>> +void *pcie_doe_get_req(DOECap *doe_cap)
>> +{
>> +    return doe_cap->write_mbox;
>> +}
>> +
>> +/* Copy the response to read mailbox buffer */
>> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp)
>> +{
>> +    uint32_t len = pcie_doe_object_len(rsp);
>> +
>> +    memcpy(doe_cap->read_mbox + doe_cap->read_mbox_len,
>> +           rsp, len * sizeof(uint32_t));
>> +    doe_cap->read_mbox_len += len;
>> +}
>> +
>> +/* Get Data Object length */
>> +uint32_t pcie_doe_object_len(void *obj)
>> +{
>> +    return (obj)? ((DOEHeader *)obj)->length : 0;
>> +}
>> +
>> +static void pcie_doe_write_mbox(DOECap *doe_cap, uint32_t val)
>> +{
>> +    memcpy(doe_cap->write_mbox + doe_cap->write_mbox_len, &val, sizeof(uint32_t));
>> +
>> +    if (doe_cap->write_mbox_len == 1) {
>> +        DOE_DBG("  Capture DOE request DO length = %d\n", val & 0x0003ffff);
>> +    }
>> +
>> +    doe_cap->write_mbox_len++;
> 
> We probably need to check that the mailbox write was full dword as the spec
> requires.  No idea what we do if it isn't as spec doesn't seem to say...
> I've queried one of our PCI experts.

Let us know their response.  If not being able to rely on full DW config access will
require a bunch of changes to accommodate.

> 
> 
>> +}
>> +
>> +static void pcie_doe_irq_assert(DOECap *doe_cap)
>> +{
>> +    PCIDevice *dev = doe_cap->doe->pdev;
>> +
>> +    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {
>> +        /* Interrupt notify */
>> +        if (msix_enabled(dev)) {
>> +            msix_notify(dev, doe_cap->cap.vec);
>> +        } else if (msi_enabled(dev)) {
>> +            msi_notify(dev, doe_cap->cap.vec);
>> +        }
>> +        /* Not support legacy IRQ */
>> +    }
>> +}
>> +
>> +static void pcie_doe_set_ready(DOECap *doe_cap, bool rdy)
>> +{
>> +    doe_cap->status.ready = rdy;
>> +
>> +    if (rdy) {
>> +        pcie_doe_irq_assert(doe_cap);
>> +    }
>> +}
>> +
>> +static void pcie_doe_set_error(DOECap *doe_cap, bool err)
>> +{
>> +    doe_cap->status.error = err;
>> +
>> +    if (err) {
>> +        pcie_doe_irq_assert(doe_cap);
>> +    }
>> +}
>> +
>> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr)
>> +{
>> +    DOECap *ptr = doe->head;
>> +
>> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
>> +    while (ptr && 
>> +           !range_covers_byte(ptr->offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
>> +        ptr = ptr->next;
>> +    }
>> +    
>> +    return ptr;
>> +}
>> +
>> +uint32_t pcie_doe_read_config(DOECap *doe_cap,
>> +                              uint32_t addr, int size)
>> +{
>> +    uint32_t ret = 0, shift, mask = 0xFFFFFFFF;
>> +    uint16_t doe_offset = doe_cap->offset;
>> +
>> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
>> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
> 
> I'd flip this around to reduce indentation + no need to be careful with alignment etc
> if we aren't returning anything.
> 
>       if (!range_cover_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)
> 		return 0;
> 
> 
>> +        addr -= doe_offset;
>> +
>> +        if (range_covers_byte(PCI_EXP_DOE_CAP, sizeof(uint32_t), addr)) {
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, INTR_SUPP,
>> +                             doe_cap->cap.intr);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM,
>> +                             doe_cap->cap.vec);
>> +        } else if (range_covers_byte(PCI_EXP_DOE_CTRL, sizeof(uint32_t), addr)) {
>> +            /* Must return ABORT=0 and GO=0 */
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_CONTROL, DOE_INTR_EN,
>> +                             doe_cap->ctrl.intr);
>> +            DOE_DBG("  CONTROL REG=%x\n", ret);
>> +        } else if (range_covers_byte(PCI_EXP_DOE_STATUS, sizeof(uint32_t), addr)) {
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_BUSY,
>> +                             doe_cap->status.busy);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_INTR_STATUS,
>> +                             doe_cap->status.intr);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DOE_ERROR,
>> +                             doe_cap->status.error);
>> +            ret = FIELD_DP32(ret, PCI_DOE_CAP_STATUS, DATA_OBJ_RDY,
>> +                             doe_cap->status.ready);
>> +        } else if (range_covers_byte(PCI_EXP_DOE_RD_DATA_MBOX, sizeof(uint32_t), addr)) {
>> +            /* Check that READY is set */
> 
> I'd clean out any comment that is obvious from the code.
> Comments get out of sync over time so there is a maintenance burden in having them such
> that we only bother if they add information.
> 
>> +            if (!doe_cap->status.ready) {
>> +                /* Return 0 if not ready */
>> +                DOE_DBG("  RD MBOX RETURN=%x\n", ret);
>> +            } else {
>> +                /* Deposit next DO DW into read mbox */
> 
> This comment is missleading.  It might not be the 'next' one. If you read
> the register twice for instance.  Much better not to have the comment
> at all on basis code is fairly obvious anyway!
> 
> As mentioned below, a read of this when nothing there is an underflow and spec
> suggests that is an error condition.
> 
>> +                DOE_DBG("  RD MBOX, DATA OBJECT READY,"
>> +                        " CURRENT DO DW OFFSET=%x\n",
>> +                        doe_cap->read_mbox_idx);
>> +
>> +                ret = doe_cap->read_mbox[doe_cap->read_mbox_idx];
>> +
>> +                DOE_DBG("  RD MBOX DW=%x\n", ret);
>> +                DOE_DBG("  RD MBOX DW OFFSET=%d, RD MBOX LENGTH=%d\n",
>> +                        doe_cap->read_mbox_idx, doe_cap->read_mbox_len);
>> +            }
>> +        } else if (range_covers_byte(PCI_EXP_DOE_WR_DATA_MBOX, sizeof(uint32_t), addr)) {
>> +            DOE_DBG("  WR MBOX, local_val =%x\n", ret);
>> +        }
>> +    }
>> +
>> +    /* Alignment */
>> +    shift = addr % sizeof(uint32_t);
>> +    if (shift) {
>> +        ret >>= shift * 8;
>> +    }
>> +    mask >>= (sizeof(uint32_t) - size) * 8;
>> +
>> +    return ret & mask;
>> +}
>> +
>> +void pcie_doe_write_config(DOECap *doe_cap,
>> +                           uint32_t addr, uint32_t val, int size)
>> +{
>> +    PCIEDOE *doe = doe_cap->doe;
>> +    uint16_t doe_offset = doe_cap->offset;
>> +    int p;
>> +    bool discard;
>> +    uint32_t shift;
>> +
>> +    DOE_DBG("  addr=%x, val=%x, size=%x\n", addr, val, size);
>> +
>> +    /* If overlaps DOE range. PCIe Capability Header won't be included. */
>> +    if (range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
> 
> As below, invert this condition and return early as it will simplify below and reduce
> indentation.
> 
>     if (!range_covers_byte(doe_offset + PCI_EXP_DOE_CAP, PCI_DOE_SIZEOF - 4, addr)) {
>          return;
>    }
> 
>> +        /* Alignment */
>> +        shift = addr % sizeof(uint32_t);
>> +        addr -= (doe_offset - shift);
>> +        val <<= shift * 8;
>> +
>> +        switch (addr) {
>> +        case PCI_EXP_DOE_CTRL:
>> +            DOE_DBG("  CONTROL=%x\n", val);
>> +            if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_ABORT)) {
>> +                /* If ABORT, clear status reg DOE Error and DOE Ready */
>> +                DOE_DBG("  Setting ABORT\n");
>> +                pcie_doe_set_ready(doe_cap, 0);
>> +                pcie_doe_set_error(doe_cap, 0);
>> +                pcie_doe_reset_mbox(doe_cap);
>> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_GO)) {
>> +                DOE_DBG("  CONTROL GO=%x\n", val);
> 
> This case is complex enough I'd factor it out as it's own function.

OK

> 
>> +                /*
>> +                 * Check protocol the incoming request in write_mbox and
>> +                 * memcpy the corresponding response to read_mbox.
>> +                 *
>> +                 * "discard" should be set up if the response callback
>> +                 * function could not deal with request (e.g. length
>> +                 * mismatch) or the protocol of request was not found.
>> +                 */
>> +                discard = DOE_DISCARD;
>> +                for (p = 0; p < doe->protocol_num; p++) {
>> +                    if (doe_cap->write_mbox[0] ==
>> +                        pcie_doe_build_protocol(&doe->protocols[p])) {
>> +                        /* Found */
>> +                        DOE_DBG("  DOE PROTOCOL=%x\n", doe_cap->write_mbox[0]);
>> +                        /*
>> +                         * Spec:
>> +                         * If the number of DW transferred does not match the
>> +                         * indicated Length for a data object, then the
>> +                         * data object must be silently discarded.
>> +                         */
>> +                        if (doe_cap->write_mbox_len ==
>> +                            pcie_doe_object_len(pcie_doe_get_req(doe_cap)))
>> +                            discard = doe->protocols[p].set_rsp(doe_cap);
>> +                        break;
>> +                    }
>> +                }
>> +
>> +                /* Set DOE Ready */
>> +                if (!discard) {
>> +                    pcie_doe_set_ready(doe_cap, 1);
>> +                } else {
>> +                    pcie_doe_reset_mbox(doe_cap);
>> +                }
>> +            } else if (FIELD_EX32(val, PCI_DOE_CAP_CONTROL, DOE_INTR_EN)) {
> 
> Spec reference needed to say why you can't write this at same time as GO.
> It may be odd thing to do, but I can't see anything saying you couldn't do this,
> for example on your very first command.
> 
>> +                doe_cap->ctrl.intr = 1;
>> +            }
>> +            break;
>> +        case PCI_EXP_DOE_RD_DATA_MBOX:
>> +            /* Read MBOX */
>> +            DOE_DBG("  INCR RD MBOX DO DW OFFSET=%d\n", doe_cap->read_mbox_idx);
>> +            doe_cap->read_mbox_idx += 1;
>> +            /* Last DW */
>> +            if (doe_cap->read_mbox_idx >= doe_cap->read_mbox_len) {
>> +                pcie_doe_reset_mbox(doe_cap);
>> +                pcie_doe_set_ready(doe_cap, 0);
>> +            }
> 
> A write of this when there is nothing there is an underflow. As is a read of this
> after the last write.  I would guess both should be error conditions.

We’ll double check and fix accordingly

> 
>> +            break;
>> +        case PCI_EXP_DOE_WR_DATA_MBOX:
>> +            /* Write MBOX */
>> +            DOE_DBG("  WR MBOX=%x, DW OFFSET = %d\n", val, doe_cap->write_mbox_len);
>> +            pcie_doe_write_mbox(doe_cap, val);
>> +            break;
>> +        case PCI_EXP_DOE_STATUS:
>> +        case PCI_EXP_DOE_CAP:
>> +        default:
>> +            break;
>> +        }
>> +    }
>> +}
>> diff --git a/include/hw/pci/pci_ids.h b/include/hw/pci/pci_ids.h
>> index 76bf3ed..636b2e8 100644
>> --- a/include/hw/pci/pci_ids.h
>> +++ b/include/hw/pci/pci_ids.h
>> @@ -157,6 +157,8 @@
>> 
>> /* Vendors and devices.  Sort key: vendor first, device next. */
>> 
>> +#define PCI_VENDOR_ID_PCI_SIG            0x0001
>> +
>> #define PCI_VENDOR_ID_LSI_LOGIC          0x1000
>> #define PCI_DEVICE_ID_LSI_53C810         0x0001
>> #define PCI_DEVICE_ID_LSI_53C895A        0x0012
>> diff --git a/include/hw/pci/pcie.h b/include/hw/pci/pcie.h
>> index 14c58eb..47d6f66 100644
>> --- a/include/hw/pci/pcie.h
>> +++ b/include/hw/pci/pcie.h
>> @@ -25,6 +25,7 @@
>> #include "hw/pci/pcie_regs.h"
>> #include "hw/pci/pcie_aer.h"
>> #include "hw/hotplug.h"
>> +#include "hw/pci/pcie_doe.h"
>> 
>> typedef enum {
>>     /* for attention and power indicator */
>> diff --git a/include/hw/pci/pcie_doe.h b/include/hw/pci/pcie_doe.h
>> new file mode 100644
>> index 0000000..f497075
>> --- /dev/null
>> +++ b/include/hw/pci/pcie_doe.h
>> @@ -0,0 +1,166 @@
>> +#ifndef PCIE_DOE_H
>> +#define PCIE_DOE_H
>> +
>> +#include "qemu/range.h"
>> +#include "qemu/typedefs.h"
>> +#include "hw/register.h"
>> +
>> +/* 
>> + * PCI DOE register defines 7.9.xx 
>> + */
>> +/* DOE Capabilities Register */
>> +#define PCI_EXP_DOE_CAP             0x04
>> +REG32(PCI_DOE_CAP_REG, 0)
>> +    FIELD(PCI_DOE_CAP_REG, INTR_SUPP, 0, 1)
>> +    FIELD(PCI_DOE_CAP_REG, DOE_INTR_MSG_NUM, 1, 11)
>> +
>> +/* DOE Control Register */
>> +#define PCI_EXP_DOE_CTRL            0x08
>> +REG32(PCI_DOE_CAP_CONTROL, 0)
>> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_ABORT, 0, 1)
>> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_INTR_EN, 1, 1)
>> +    FIELD(PCI_DOE_CAP_CONTROL, DOE_GO, 31, 1)
>> +
>> +/* DOE Status Register  */
>> +#define PCI_EXP_DOE_STATUS          0x0c
>> +REG32(PCI_DOE_CAP_STATUS, 0)
>> +    FIELD(PCI_DOE_CAP_STATUS, DOE_BUSY, 0, 1)
>> +    FIELD(PCI_DOE_CAP_STATUS, DOE_INTR_STATUS, 1, 1)
>> +    FIELD(PCI_DOE_CAP_STATUS, DOE_ERROR, 2, 1)
>> +    FIELD(PCI_DOE_CAP_STATUS, DATA_OBJ_RDY, 31, 1)
>> +
>> +/* DOE Write Data Mailbox Register  */
>> +#define PCI_EXP_DOE_WR_DATA_MBOX    0x10
>> +
>> +/* DOE Read Data Mailbox Register  */
>> +#define PCI_EXP_DOE_RD_DATA_MBOX    0x14
>> +
>> +/* 
>> + * PCI DOE register defines 7.9.xx 
>> + */
>> +/* Table 7-x2 */
>> +#define PCI_SIG_DOE_DISCOVERY       0x00
>> +#define PCI_SIG_DOE_CMA             0x01
>> +
>> +#define DATA_OBJ_BUILD_HEADER1(v, p)  ((p << 16) | v)
>> +
>> +/* Figure 6-x1 */
>> +#define DATA_OBJECT_HEADER1_OFFSET  0
>> +#define DATA_OBJECT_HEADER2_OFFSET  1
>> +#define DATA_OBJECT_CONTENT_OFFSET  2
>> +
>> +#define PCI_DOE_MAX_DW_SIZE (1 << 18)
>> +#define PCI_DOE_PROTOCOL_MAX 256
>> +
>> +/*
>> + * Self-defined Marco
>> + */
>> +/* Request/Response State */
>> +#define DOE_SUCCESS 0
>> +#define DOE_DISCARD 1
> Drop these. As mentioned in previous review.  These are just
> obvious bools.
> 
>> +
>> +/* An invalid vector number for MSI/MSI-x */
>> +#define DOE_NO_INTR (~0)
>> +
>> +/*
>> + * DOE Protocol - Data Object
>> + */
>> +typedef struct DOEHeader DOEHeader;
>> +typedef struct DOEProtocol DOEProtocol;
>> +typedef struct PCIEDOE PCIEDOE;
>> +typedef struct DOECap DOECap;
>> +
>> +struct DOEHeader {
>> +    uint16_t vendor_id;
>> +    uint8_t doe_type;
>> +    uint8_t reserved;
>> +    struct {
>> +        uint32_t length:18;
>> +        uint32_t reserved2:14;
> 
> As before, bitfields are not a good idea in this sort of code in general due to lack
> of standard handling in the C spec.
> 
>> +    };
>> +} QEMU_PACKED;
>> +
>> +/* Protocol infos and rsp function callback */
>> +struct DOEProtocol {
>> +    uint16_t vendor_id;
>> +    uint8_t doe_type;
>> +
>> +    bool (*set_rsp)(DOECap *);
>> +};
>> +
>> +struct PCIEDOE {
>> +    PCIDevice *pdev;
> 
> Having the PCIDevice in here only allows you to drop a parameter in read and write
> calls. I'd not bother given the nest of pointers you have as a result.
> Just pass both the PCIDOE and the PCIDevice into those two functions.
> 
>> +    DOECap *head;
> 
> I can sort of get what you are doing with this list of DOEs but to mind it's the wrong
> abstraction as it doesn't fit how I think these will be used.
> 
> It's not a case that there will be N DOE mailboxes each of which supports
> all the same protocols.  I suspect quite the opposite.
> 
> Each DOE may support multiple protocols but the most obvious reason you'd
> do multiple DOEs is because they support different protocols.

Agree

> 
> If we want to support the same protocol on mulitple DOE instances then we'd
> register it for each of them.
> 
> Thus I'd drop the list handling and instead  have for example
> 
> struct cxl_type3_dev {
> 
> ...
> 
>    PCIDOE doe_ima;
>    PCIDOE doe_cdat;
> };
> 
> In the read and write calls then just call pcie_doe_xxxx_config() twice.
> You do have to handle the miss vs hit stuff though (similar problem that
> lead to ugly code in my version).

OK we’ll handle it better in next patch.

> 
> 
>> +
>> +    /* Protocols and its callback response */
>> +    DOEProtocol protocols[PCI_DOE_PROTOCOL_MAX];
>> +    uint32_t protocol_num;
>> +};
>> +
>> +struct DOECap {
>> +    PCIEDOE *doe;
> 
> Following suggestion to get rid of the list, you should also then combine
> PCIEDOE and the DOECap as they match 1 to 1.
> 
>> +
>> +    /* Capability offset */
>> +    uint16_t offset;
>> +
>> +    /* Next DOE capability */
>> +    DOECap *next;
>> +
>> +    /* Capability */
>> +    struct {
>> +        bool intr;
>> +        uint16_t vec;
>> +    } cap;
>> +
>> +    /* Control */
>> +    struct {
>> +        bool abort;
>> +        bool intr;
>> +        bool go;
>> +    } ctrl;
>> +
>> +    /* Status */
>> +    struct {
>> +        bool busy;
>> +        bool intr;
>> +        bool error;
>> +        bool ready;
>> +    } status;
>> +
>> +    /* Mailbox buffer for device */
>> +    uint32_t *write_mbox;
>> +    uint32_t *read_mbox;
>> +
>> +    /* Mailbox position indicator */
>> +    uint32_t read_mbox_idx;
>> +    uint32_t read_mbox_len;
>> +    uint32_t write_mbox_len;
>> +};
>> +
>> +void pcie_doe_init(PCIDevice *dev, PCIEDOE *doe);
>> +int pcie_cap_doe_add(PCIEDOE *doe, uint16_t offset, bool intr, uint16_t vec);
>> +void pcie_doe_uninit(PCIEDOE *doe);
>> +void pcie_doe_register_protocol(PCIEDOE *doe, uint16_t vendor_id,
>> +                                uint8_t doe_type,
>> +                                bool (*set_rsp)(DOECap *));
>> +uint32_t pcie_doe_build_protocol(DOEProtocol *p);
>> +DOECap *pcie_doe_covers_addr(PCIEDOE *doe, uint32_t addr);
>> +uint32_t pcie_doe_read_config(DOECap *doe_cap, uint32_t addr, int size);
>> +void pcie_doe_write_config(DOECap *doe_cap, uint32_t addr,
>> +                           uint32_t val, int size);
>> +
>> +/* Utility functions for set_rsp in DOEProtocol */
>> +void *pcie_doe_get_req(DOECap *doe_cap);
>> +void pcie_doe_set_rsp(DOECap *doe_cap, void *rsp);
>> +uint32_t pcie_doe_object_len(void *obj);
>> +
>> +#define DOE_DEBUG
>> +#ifdef DOE_DEBUG
>> +#define DOE_DBG(...) printf(__VA_ARGS__)
>> +#else
>> +#define DOE_DBG(...) {}
>> +#endif
> 
> Tidy this stuff up (i.e. get rid of it) for next version.  It's fine when
> you are doing early debug, but we don't want it in the version for review.

Got it!

> 
>> +
>> +#define dwsizeof(s) ((sizeof(s) + 4 - 1) / 4)
> 
> Not used in this patch.  Move it down towards code that uses it or at very 
> least only introduce it when used.
> 
>> +
>> +#endif /* PCIE_DOE_H */
>> diff --git a/include/hw/pci/pcie_regs.h b/include/hw/pci/pcie_regs.h
>> index 1db86b0..963dc2e 100644
>> --- a/include/hw/pci/pcie_regs.h
>> +++ b/include/hw/pci/pcie_regs.h
>> @@ -179,4 +179,8 @@ typedef enum PCIExpLinkWidth {
>> #define PCI_ACS_VER                     0x1
>> #define PCI_ACS_SIZEOF                  8
>> 
>> +/* DOE Capability Register Fields */
>> +#define PCI_DOE_VER                     0x1
>> +#define PCI_DOE_SIZEOF                  24
>> +
>> #endif /* QEMU_PCIE_REGS_H */
>> diff --git a/include/standard-headers/linux/pci_regs.h b/include/standard-headers/linux/pci_regs.h
>> index e709ae8..4a7b7a4 100644
>> --- a/include/standard-headers/linux/pci_regs.h
>> +++ b/include/standard-headers/linux/pci_regs.h
> 
> Pull this change out as a separate patch.  This header gets fetched via a script
> so we will only find this here if we update the source file in the kernel.

Got it!

> 
> That should happen soon anyway as we add the support to read this.
> 
> 
>> @@ -730,7 +730,8 @@
>> #define PCI_EXT_CAP_ID_DVSEC	0x23	/* Designated Vendor-Specific */
>> #define PCI_EXT_CAP_ID_DLF	0x25	/* Data Link Feature */
>> #define PCI_EXT_CAP_ID_PL_16GT	0x26	/* Physical Layer 16.0 GT/s */
>> -#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_PL_16GT
>> +#define PCI_EXT_CAP_ID_DOE      0x2E    /* Data Object Exchange */
>> +#define PCI_EXT_CAP_ID_MAX	PCI_EXT_CAP_ID_DOE
>> 
>> #define PCI_EXT_CAP_DSN_SIZEOF	12
>> #define PCI_EXT_CAP_MCAST_ENDPOINT_SIZEOF 40
> 

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Jonathan Cameron]

On Fri, 12 Feb 2021 16:58:21 -0500
Chris Browy <cbrowy@avery-design.com> wrote:

> > On Feb 12, 2021, at 11:24 AM, Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:
> > 
> > On Tue, 9 Feb 2021 15:35:49 -0500
> > Chris Browy <cbrowy@avery-design.com> wrote:
> > 
> > Run ./scripts/checkpatch.pl over the patches and fix all the warnings before
> > posting.  It will save time by clearing out most of the minor formatting issues
> > and similar that inevitably sneak in during development.
> >   
> Excellent suggestion.  We’re still newbies!
> 
> > The biggest issue I'm seeing in here is that the abstraction of
> > multiple DOE capabiltiies accessing same protocols doesn't make sense.
> > 
> > Each DOE ecap region and hence mailbox can have it's own set of
> > (possibly  overlapping) protocols.
> > 
> > From the ECN:
> > "It is permitted for a protocol using data object exchanges to require
> > that a Function implement a unique instance of DOE for that specific
> > protocol, and/or to allow sharing of a DOE instance to only a specific
> > set of protocols using data object exchange, and/or to allow a Function
> > to implement multiple instances of DOE supporting the specific protocol."
> > 
> > Tightly couple the ECAP and DOE.  If we are in the multiple instances
> > of DOE supporting a specific protocol case, then register it separately
> > for each one.  The individual device emulation then needs to deal with
> > any possible clashes etc.  
> 
> Not sure how configurable we want to make the device.  It is a simple type 3
> device after all. 

Agreed, but what I (or someone else) really doesn't want to have to do
in the future is reimplement DOE because we made design decisions that make
this version hard to reuse.  Unless it is particularly nasty to do we should
try to design something that is generally useful rather than targeted to
closely at the specific case we are dealing with.

I'd argue the ECAP and the DOE mailbox are always tightly coupled 1-to-1.
Whether the device wants to implement multiple protocols on each DOE mailbox
or indeed run individual protocols on multiple DOE mailboxes is a design
decision, but the actual mechanics of DOE match up with the config
space structures anything else is impdef on the device.

> 
> The DOE spec does leave it pretty arbitrary regarding N DOE instances (DOE 
> Extended Cap entry points) for M protocols, including where N>1 and M=1.  
> Currently we implement N=2 DOE caps (instances), one for CDAT, one for 
> Compliance Mode.[
> 
> Maybe a more complex MLD device might have one or more DOE instances 
> for the CDAT protocol alone to define each HDM but currently we only have 
> one pmem (SLD) so we can’t really do much more than what’s supported.
> 
> Open to further suggestion though.  Based on answer to above we’ll follow 
> the suggestion lower in the code review regarding 
> 
...

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Chris Browy]

> On Feb 18, 2021, at 2:11 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
> 
> On Fri, 12 Feb 2021 16:58:21 -0500
> Chris Browy <cbrowy@avery-design.com> wrote:
> 
>>> On Feb 12, 2021, at 11:24 AM, Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:
>>> 
>>> On Tue, 9 Feb 2021 15:35:49 -0500
>>> Chris Browy <cbrowy@avery-design.com> wrote:
>>> 
>>> Run ./scripts/checkpatch.pl over the patches and fix all the warnings before
>>> posting.  It will save time by clearing out most of the minor formatting issues
>>> and similar that inevitably sneak in during development.
>>> 
>> Excellent suggestion.  We’re still newbies!
>> 
>>> The biggest issue I'm seeing in here is that the abstraction of
>>> multiple DOE capabiltiies accessing same protocols doesn't make sense.
>>> 
>>> Each DOE ecap region and hence mailbox can have it's own set of
>>> (possibly  overlapping) protocols.
>>> 
>>> From the ECN:
>>> "It is permitted for a protocol using data object exchanges to require
>>> that a Function implement a unique instance of DOE for that specific
>>> protocol, and/or to allow sharing of a DOE instance to only a specific
>>> set of protocols using data object exchange, and/or to allow a Function
>>> to implement multiple instances of DOE supporting the specific protocol."
>>> 
>>> Tightly couple the ECAP and DOE.  If we are in the multiple instances
>>> of DOE supporting a specific protocol case, then register it separately
>>> for each one.  The individual device emulation then needs to deal with
>>> any possible clashes etc.  
>> 
>> Not sure how configurable we want to make the device.  It is a simple type 3
>> device after all. 
> 
> Agreed, but what I (or someone else) really doesn't want to have to do
> in the future is reimplement DOE because we made design decisions that make
> this version hard to reuse.  Unless it is particularly nasty to do we should
> try to design something that is generally useful rather than targeted to
> closely at the specific case we are dealing with.
> 
> I'd argue the ECAP and the DOE mailbox are always tightly coupled 1-to-1.
> Whether the device wants to implement multiple protocols on each DOE mailbox
> or indeed run individual protocols on multiple DOE mailboxes is a design
> decision, but the actual mechanics of DOE match up with the config
> space structures anything else is impdef on the device.

Yes I agree that there is 1-to-1 between DOE extended cap (ECAP) and DOE
Mailbox.  If we want to provide complete flexibility we should let the user pass 
device property arrays to QEMU command for how many DOE ECAP’s to build 
out and how to assign protocol(s) to each of them.  Array index is the DOE 
instance #.

Also we can provide a property for cdat binary (blob) filename to initialize 
the CDAT structure[entries].  This just reads in whatever mix of CDAT structure
types are in the blob.

-device cxl-type3,bus=rp0,memdev=cxl-mem1,id=cxl-pmem0,size=256M \
    doe-ecap-instances=2 \
    doe-ecap[0]=5 // bitwise OR for protocols shared
    doe-ecap[1]=2 //bitwise OR for protocols shared
    doe-ecap-cdat[1]=mycdat.bin

where let’s say protocols bitvector
bit [0]=CMA
bit [1]=CDAT
bit [2}=Compliance

Let me know if you some better alternatives and we’ll implement it.


> 
>> 
>> The DOE spec does leave it pretty arbitrary regarding N DOE instances (DOE 
>> Extended Cap entry points) for M protocols, including where N>1 and M=1.  
>> Currently we implement N=2 DOE caps (instances), one for CDAT, one for 
>> Compliance Mode.[
>> 
>> Maybe a more complex MLD device might have one or more DOE instances 
>> for the CDAT protocol alone to define each HDM but currently we only have 
>> one pmem (SLD) so we can’t really do much more than what’s supported.
>> 
>> Open to further suggestion though.  Based on answer to above we’ll follow 
>> the suggestion lower in the code review regarding 
>> 
> ...
> 

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Jonathan Cameron]

On Thu, 18 Feb 2021 19:46:54 -0500
Chris Browy <cbrowy@avery-design.com> wrote:

> > On Feb 18, 2021, at 2:11 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
> > 
> > On Fri, 12 Feb 2021 16:58:21 -0500
> > Chris Browy <cbrowy@avery-design.com> wrote:
> >   
> >>> On Feb 12, 2021, at 11:24 AM, Jonathan Cameron <Jonathan.Cameron@Huawei.com> wrote:
> >>> 
> >>> On Tue, 9 Feb 2021 15:35:49 -0500
> >>> Chris Browy <cbrowy@avery-design.com> wrote:
> >>> 
> >>> Run ./scripts/checkpatch.pl over the patches and fix all the warnings before
> >>> posting.  It will save time by clearing out most of the minor formatting issues
> >>> and similar that inevitably sneak in during development.
> >>>   
> >> Excellent suggestion.  We’re still newbies!
> >>   
> >>> The biggest issue I'm seeing in here is that the abstraction of
> >>> multiple DOE capabiltiies accessing same protocols doesn't make sense.
> >>> 
> >>> Each DOE ecap region and hence mailbox can have it's own set of
> >>> (possibly  overlapping) protocols.
> >>> 
> >>> From the ECN:
> >>> "It is permitted for a protocol using data object exchanges to require
> >>> that a Function implement a unique instance of DOE for that specific
> >>> protocol, and/or to allow sharing of a DOE instance to only a specific
> >>> set of protocols using data object exchange, and/or to allow a Function
> >>> to implement multiple instances of DOE supporting the specific protocol."
> >>> 
> >>> Tightly couple the ECAP and DOE.  If we are in the multiple instances
> >>> of DOE supporting a specific protocol case, then register it separately
> >>> for each one.  The individual device emulation then needs to deal with
> >>> any possible clashes etc.    
> >> 
> >> Not sure how configurable we want to make the device.  It is a simple type 3
> >> device after all.   
> > 
> > Agreed, but what I (or someone else) really doesn't want to have to do
> > in the future is reimplement DOE because we made design decisions that make
> > this version hard to reuse.  Unless it is particularly nasty to do we should
> > try to design something that is generally useful rather than targeted to
> > closely at the specific case we are dealing with.
> > 
> > I'd argue the ECAP and the DOE mailbox are always tightly coupled 1-to-1.
> > Whether the device wants to implement multiple protocols on each DOE mailbox
> > or indeed run individual protocols on multiple DOE mailboxes is a design
> > decision, but the actual mechanics of DOE match up with the config
> > space structures anything else is impdef on the device.  
> 
> Yes I agree that there is 1-to-1 between DOE extended cap (ECAP) and DOE
> Mailbox.  If we want to provide complete flexibility we should let the user pass 
> device property arrays to QEMU command for how many DOE ECAP’s to build 
> out and how to assign protocol(s) to each of them.  Array index is the DOE 
> instance #.
> 
> Also we can provide a property for cdat binary (blob) filename to initialize 
> the CDAT structure[entries].  This just reads in whatever mix of CDAT structure
> types are in the blob.
> 
> -device cxl-type3,bus=rp0,memdev=cxl-mem1,id=cxl-pmem0,size=256M \
>     doe-ecap-instances=2 \
>     doe-ecap[0]=5 // bitwise OR for protocols shared
>     doe-ecap[1]=2 //bitwise OR for protocols shared
>     doe-ecap-cdat[1]=mycdat.bin
> 
> where let’s say protocols bitvector
> bit [0]=CMA
> bit [1]=CDAT
> bit [2}=Compliance
> 
> Let me know if you some better alternatives and we’ll implement it.
> 

Gut feeling for DOE is that the particular combination of protocol and
ECAP/DOE is device dependent. As such...

I'm not sure we actually want to expose it as command line controllable at all.
If we do, I'd suggest a small number of sane choices that exercise cases
we want to check.

From a testing point of view, 2 DOE, one of which supports multiple
protocols and we will have enough to test likely failure modes in the code.

The one protocol running on multiple mailboxes is already covered by the
discovery protocol which they all support.  That might not exercise
all the potential problems on the emulator side (as no need to do
locking etc) but it will proabbly exercise those in the OS and firmware.

Almost all users of DOE functionality in QEMU in the long run are likely
to be emulating a particular device so will hard code the DOE instances present
on that device in their emulation of whatever PCIe device they are
emulating.

This is no different to picking a particular layout for config space.
We could make it fully flexible, but it's rarely useful to do so.

If anyone wants to check something unusual, they can hack it into
QEMU.

As a side note, a protocol bit vector is going to unmaintainable as
there will be lots of protocols and last thing we want is that vector
to mean different things on different emulated PCI devices.

Jonathan



> 
> >   
> >> 
> >> The DOE spec does leave it pretty arbitrary regarding N DOE instances (DOE 
> >> Extended Cap entry points) for M protocols, including where N>1 and M=1.  
> >> Currently we implement N=2 DOE caps (instances), one for CDAT, one for 
> >> Compliance Mode.[
> >> 
> >> Maybe a more complex MLD device might have one or more DOE instances 
> >> for the CDAT protocol alone to define each HDM but currently we only have 
> >> one pmem (SLD) so we can’t really do much more than what’s supported.
> >> 
> >> Open to further suggestion though.  Based on answer to above we’ll follow 
> >> the suggestion lower in the code review regarding 
> >>   
> > ...
> >   
> 

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Jonathan Cameron]

On Tue, 9 Feb 2021 15:35:49 -0500
Chris Browy <cbrowy@avery-design.com> wrote:

Hi Chris,

One more thing hit whilst debugging linux side of this.

> +static void pcie_doe_irq_assert(DOECap *doe_cap)
> +{
> +    PCIDevice *dev = doe_cap->doe->pdev;
> +
> +    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {


need something like

doe_cap->status.intr = 1;

I think or anyone checking the status register is going to think
this interrupt is spurious.

Otherwise all seems to work. I need to do a bit of tidying up on
kernel code but should be able to send out early next week.

> +        /* Interrupt notify */
> +        if (msix_enabled(dev)) {
> +            msix_notify(dev, doe_cap->cap.vec);
> +        } else if (msi_enabled(dev)) {
> +            msi_notify(dev, doe_cap->cap.vec);
> +        }
> +        /* Not support legacy IRQ */
> +    }
> +}

Re: [RFC PATCH v2 1/2] Basic PCIe DOE support

[Chris Browy]

> On Mar 4, 2021, at 2:21 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
> 
> On Tue, 9 Feb 2021 15:35:49 -0500
> Chris Browy <cbrowy@avery-design.com> wrote:
> 
> Hi Chris,
> 
> One more thing hit whilst debugging linux side of this.
> 
>> +static void pcie_doe_irq_assert(DOECap *doe_cap)
>> +{
>> +    PCIDevice *dev = doe_cap->doe->pdev;
>> +
>> +    if (doe_cap->cap.intr && doe_cap->ctrl.intr) {
> 
> 
> need something like
> 
> doe_cap->status.intr = 1;
> 
> I think or anyone checking the status register is going to think
> this interrupt is spurious.

You’re absolutely right, good catch!

> 
> Otherwise all seems to work. I need to do a bit of tidying up on
> kernel code but should be able to send out early next week.
> 

We’re putting out the v3 by end of this week.  We’re spent a bit longer
tidying up on our end but sounds like coming together real soon in 5.12 
release!

>> +        /* Interrupt notify */
>> +        if (msix_enabled(dev)) {
>> +            msix_notify(dev, doe_cap->cap.vec);
>> +        } else if (msi_enabled(dev)) {
>> +            msi_notify(dev, doe_cap->cap.vec);
>> +        }
>> +        /* Not support legacy IRQ */
>> +    }
>> +}

[RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Chris Browy]

---
 hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
 hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
 include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
 include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
 include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
 include/hw/cxl/cxl_device.h    |   3 +
 include/hw/cxl/cxl_pci.h       |   4 +
 7 files changed, 846 insertions(+)
 create mode 100644 include/hw/cxl/cxl_cdat.h
 create mode 100644 include/hw/cxl/cxl_compl.h

diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
index e1bcee5..fc6c538 100644
--- a/hw/cxl/cxl-component-utils.c
+++ b/hw/cxl/cxl-component-utils.c
@@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
     range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
     cxl->dvsec_offset += length;
 }
+
+/* Return the sum of bytes */
+static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
+{
+    cs->base = base;
+    cs->length = len;
+}
+
+void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
+{
+    uint8_t sum = 0;
+    uint32_t len = 0;
+    int i, j;
+
+    cxl_cstate->cdat_ent_len = 7;
+    cxl_cstate->cdat_ent =
+        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
+
+    cdat_ent_init(&cxl_cstate->cdat_ent[0],
+                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
+    cdat_ent_init(&cxl_cstate->cdat_ent[1],
+                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
+    cdat_ent_init(&cxl_cstate->cdat_ent[2],
+                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
+    cdat_ent_init(&cxl_cstate->cdat_ent[3],
+                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
+    cdat_ent_init(&cxl_cstate->cdat_ent[4],
+                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
+    cdat_ent_init(&cxl_cstate->cdat_ent[5],
+                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
+    cdat_ent_init(&cxl_cstate->cdat_ent[6],
+                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
+
+    /* Set the DSMAS entry, ent = 1 */
+    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
+    cxl_cstate->dsmas.header.reserved = 0x0;
+    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
+    cxl_cstate->dsmas.DSMADhandle = 0x0;
+    cxl_cstate->dsmas.flags = 0x0;
+    cxl_cstate->dsmas.reserved2 = 0x0;
+    cxl_cstate->dsmas.DPA_base = 0x0;
+    cxl_cstate->dsmas.DPA_length = 0x40000;
+
+    /* Set the DSLBIS entry, ent = 2 */
+    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
+    cxl_cstate->dslbis.header.reserved = 0;
+    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
+    cxl_cstate->dslbis.handle = 0;
+    cxl_cstate->dslbis.flags = 0;
+    cxl_cstate->dslbis.data_type = 0;
+    cxl_cstate->dslbis.reserved2 = 0;
+    cxl_cstate->dslbis.entry_base_unit = 0;
+    cxl_cstate->dslbis.entry[0] = 0;
+    cxl_cstate->dslbis.entry[1] = 0;
+    cxl_cstate->dslbis.entry[2] = 0;
+    cxl_cstate->dslbis.reserved3 = 0;
+
+    /* Set the DSMSCIS entry, ent = 3 */
+    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
+    cxl_cstate->dsmscis.header.reserved = 0;
+    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
+    cxl_cstate->dsmscis.DSMASH_handle = 0;
+    cxl_cstate->dsmscis.reserved2[0] = 0;
+    cxl_cstate->dsmscis.reserved2[1] = 0;
+    cxl_cstate->dsmscis.reserved2[2] = 0;
+    cxl_cstate->dsmscis.memory_side_cache_size = 0;
+    cxl_cstate->dsmscis.cache_attributes = 0;
+
+    /* Set the DSIS entry, ent = 4 */
+    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
+    cxl_cstate->dsis.header.reserved = 0;
+    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
+    cxl_cstate->dsis.flags = 0;
+    cxl_cstate->dsis.handle = 0;
+    cxl_cstate->dsis.reserved2 = 0;
+
+    /* Set the DSEMTS entry, ent = 5 */
+    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
+    cxl_cstate->dsemts.header.reserved = 0;
+    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
+    cxl_cstate->dsemts.DSMAS_handle = 0;
+    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
+    cxl_cstate->dsemts.reserved2 = 0;
+    cxl_cstate->dsemts.DPA_offset = 0;
+    cxl_cstate->dsemts.DPA_length = 0;
+
+    /* Set the SSLBIS entry, ent = 6 */
+    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
+    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
+    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
+    cxl_cstate->sslbis.sslbis_h.data_type = 0;
+    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
+    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
+    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
+    /* Set the SSLBE entry */
+    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
+    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
+    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
+    cxl_cstate->sslbis.sslbe[0].reserved = 0;
+    /* Set the SSLBE entry */
+    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
+    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
+    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
+    cxl_cstate->sslbis.sslbe[1].reserved = 0;
+    /* Set the SSLBE entry */
+    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
+    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
+    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
+    cxl_cstate->sslbis.sslbe[2].reserved = 0;
+
+    /* Set CDAT header, ent = 0 */
+    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
+    cxl_cstate->cdat_header.reserved[0] = 0;
+    cxl_cstate->cdat_header.reserved[1] = 0;
+    cxl_cstate->cdat_header.reserved[2] = 0;
+    cxl_cstate->cdat_header.reserved[3] = 0;
+    cxl_cstate->cdat_header.reserved[4] = 0;
+    cxl_cstate->cdat_header.reserved[5] = 0;
+    cxl_cstate->cdat_header.sequence = 0;
+
+    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
+        /* Add length of each CDAT struct to total length */
+        len = cxl_cstate->cdat_ent[i].length;
+        cxl_cstate->cdat_header.length += len;
+
+        /* Calculate checksum of each CDAT struct */
+        for (j = 0; j < len; j++) {
+            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
+        }
+    }
+    cxl_cstate->cdat_header.checksum = ~sum + 1;
+}
diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
index d091e64..86c762f 100644
--- a/hw/mem/cxl_type3.c
+++ b/hw/mem/cxl_type3.c
@@ -13,6 +13,150 @@
 #include "qemu/rcu.h"
 #include "sysemu/hostmem.h"
 #include "hw/cxl/cxl.h"
+#include "hw/pci/msi.h"
+#include "hw/pci/msix.h"
+
+uint32_t cxl_doe_compliance_init(DOECap *doe_cap)
+{
+    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
+    uint32_t req;
+    uint32_t byte_cnt = 0;
+
+    DOE_DBG(">> %s\n",  __func__);
+
+    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
+        ->req_code;
+    switch (req) {
+    case CXL_COMP_MODE_CAP:
+        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);
+        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
+        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
+        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
+        cxl_cstate->doe_resp.cap_rsp.header.length =
+            dwsizeof(struct cxl_compliance_mode_cap_rsp);
+        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
+        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
+        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
+        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
+        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
+        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
+        break;
+    case CXL_COMP_MODE_STATUS:
+        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
+        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
+        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
+        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
+        cxl_cstate->doe_resp.status_rsp.header.length =
+            dwsizeof(struct cxl_compliance_mode_status_rsp);
+        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
+        cxl_cstate->doe_resp.status_rsp.version = 0x1;
+        cxl_cstate->doe_resp.status_rsp.length = 0x14;
+        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
+        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
+        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
+        break;
+    default:
+        break;
+    }
+
+    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
+    DOE_DBG("<< %s\n",  __func__);
+    return byte_cnt;
+}
+
+
+bool cxl_doe_compliance_rsp(DOECap *doe_cap)
+{
+    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
+    uint32_t byte_cnt;
+    uint32_t dw_cnt;
+
+    DOE_DBG(">> %s\n",  __func__);
+
+    byte_cnt = cxl_doe_compliance_init(doe_cap);
+    dw_cnt = byte_cnt / 4;
+    memcpy(doe_cap->read_mbox,
+           cxl_cstate->doe_resp.data_byte, byte_cnt);
+
+    doe_cap->read_mbox_len += dw_cnt;
+
+    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
+            doe_cap->read_mbox_len,
+            *(doe_cap->read_mbox + dw_cnt - 1));
+
+    DOE_DBG("<< %s\n",  __func__);
+
+    return DOE_SUCCESS;
+}
+
+bool cxl_doe_cdat_rsp(DOECap *doe_cap)
+{
+    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
+    uint16_t ent;
+    void *base;
+    uint32_t len;
+    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
+
+    ent = req->entry_handle;
+    base = cxl_cstate->cdat_ent[ent].base;
+    len = cxl_cstate->cdat_ent[ent].length;
+
+    struct cxl_cdat_rsp rsp = {
+        .header = {
+            .vendor_id = CXL_VENDOR_ID,
+            .doe_type = CXL_DOE_TABLE_ACCESS,
+            .reserved = 0x0,
+            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
+        },
+        .req_code = CXL_DOE_TAB_RSP,
+        .table_type = CXL_DOE_TAB_TYPE_CDAT,
+        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
+    };
+
+    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
+    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
+
+    doe_cap->read_mbox_len += rsp.header.length;
+    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
+
+    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
+        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
+    }
+
+    return DOE_SUCCESS;
+}
+
+static uint32_t ct3d_config_read(PCIDevice *pci_dev,
+                            uint32_t addr, int size)
+{
+    CXLType3Dev *ct3d = CT3(pci_dev);
+    PCIEDOE *doe = &ct3d->doe;
+    DOECap *doe_cap;
+
+    doe_cap = pcie_doe_covers_addr(doe, addr);
+    if (doe_cap) {
+        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
+        return pcie_doe_read_config(doe_cap, addr, size);
+    } else {
+        return pci_default_read_config(pci_dev, addr, size);
+    }
+}
+
+static void ct3d_config_write(PCIDevice *pci_dev,
+                            uint32_t addr, uint32_t val, int size)
+{
+    CXLType3Dev *ct3d = CT3(pci_dev);
+    PCIEDOE *doe = &ct3d->doe;
+    DOECap *doe_cap;
+
+    doe_cap = pcie_doe_covers_addr(doe, addr);
+    if (doe_cap) {
+        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
+        pcie_doe_write_config(doe_cap, addr, val, size);
+    } else {
+        pci_default_write_config(pci_dev, addr, val, size);
+    }
+}
 
 static void build_dvsecs(CXLType3Dev *ct3d)
 {
@@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
     ComponentRegisters *regs = &cxl_cstate->crb;
     MemoryRegion *mr = &regs->component_registers;
     uint8_t *pci_conf = pci_dev->config;
+    unsigned short msix_num = 2;
+    int rc;
+    int i;
 
     if (!ct3d->cxl_dstate.pmem) {
         cxl_setup_memory(ct3d, errp);
@@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
                      PCI_BASE_ADDRESS_SPACE_MEMORY |
                          PCI_BASE_ADDRESS_MEM_TYPE_64,
                      &ct3d->cxl_dstate.device_registers);
+
+    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
+    for (i = 0; i < msix_num; i++) {
+        msix_vector_use(pci_dev, i);
+    }
+
+    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */
+
+    pcie_doe_init(pci_dev, &ct3d->doe);
+    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);
+    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
+    if (rc) {
+        error_setg(errp, "fail to add DOE cap");
+        return;
+    }
+
+    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
+                               cxl_doe_compliance_rsp);
+    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
+                               cxl_doe_cdat_rsp);
+
+    cxl_doe_cdat_init(cxl_cstate);
 }
 
 static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
@@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
     DeviceClass *dc = DEVICE_CLASS(oc);
     PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
     MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
+
+    pc->config_write = ct3d_config_write;
+    pc->config_read = ct3d_config_read;
     CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
 
     pc->realize = ct3_realize;
diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
new file mode 100644
index 0000000..fbbd494
--- /dev/null
+++ b/include/hw/cxl/cxl_cdat.h
@@ -0,0 +1,120 @@
+#include "hw/cxl/cxl_pci.h"
+
+
+enum {
+    CXL_DOE_COMPLIANCE             = 0,
+    CXL_DOE_TABLE_ACCESS           = 2,
+    CXL_DOE_MAX_PROTOCOL
+};
+
+#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
+#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
+
+/*
+ * DOE CDAT Table Protocol (CXL Spec)
+ */
+#define CXL_DOE_TAB_REQ 0
+#define CXL_DOE_TAB_RSP 0
+#define CXL_DOE_TAB_TYPE_CDAT 0
+#define CXL_DOE_TAB_ENT_MAX 0xFFFF
+
+/* Read Entry Request, 8.1.11.1 Table 134 */
+struct cxl_cdat {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t table_type;
+    uint16_t entry_handle;
+} QEMU_PACKED;
+
+/* Read Entry Response, 8.1.11.1 Table 135 */
+#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
+
+/*
+ * CDAT Table Structure (CDAT Spec)
+ */
+#define CXL_CDAT_REV 1
+
+/* Data object header */
+struct cdat_table_header {
+    uint32_t length;    /* Length of table in bytes, including this header */
+    uint8_t revision;   /* ACPI Specification minor version number */
+    uint8_t checksum;   /* To make sum of entire table == 0 */
+    uint8_t reserved[6];
+    uint32_t sequence;  /* ASCII table signature */
+} QEMU_PACKED;
+
+/* Values for subtable type in CDAT structures */
+enum cdat_type {
+    CDAT_TYPE_DSMAS = 0,
+    CDAT_TYPE_DSLBIS = 1,
+    CDAT_TYPE_DSMSCIS = 2,
+    CDAT_TYPE_DSIS = 3,
+    CDAT_TYPE_DSEMTS = 4,
+    CDAT_TYPE_SSLBIS = 5,
+    CDAT_TYPE_MAX
+};
+
+struct cdat_sub_header {
+    uint8_t type;
+    uint8_t reserved;
+    uint16_t length;
+};
+
+/* CDAT Structure Subtables */
+struct cdat_dsmas {
+    struct cdat_sub_header header;
+    uint8_t DSMADhandle;
+    uint8_t flags;
+    uint16_t reserved2;
+    uint64_t DPA_base;
+    uint64_t DPA_length;
+} QEMU_PACKED;
+
+struct cdat_dslbis {
+    struct cdat_sub_header header;
+    uint8_t handle;
+    uint8_t flags;
+    uint8_t data_type;
+    uint8_t reserved2;
+    uint64_t entry_base_unit;
+    uint16_t entry[3];
+    uint16_t reserved3;
+} QEMU_PACKED;
+
+struct cdat_dsmscis {
+    struct cdat_sub_header header;
+    uint8_t DSMASH_handle;
+    uint8_t reserved2[3];
+    uint64_t memory_side_cache_size;
+    uint32_t cache_attributes;
+} QEMU_PACKED;
+
+struct cdat_dsis {
+    struct cdat_sub_header header;
+    uint8_t flags;
+    uint8_t handle;
+    uint16_t reserved2;
+} QEMU_PACKED;
+
+struct cdat_dsemts {
+    struct cdat_sub_header header;
+    uint8_t DSMAS_handle;
+    uint8_t EFI_memory_type_attr;
+    uint16_t reserved2;
+    uint64_t DPA_offset;
+    uint64_t DPA_length;
+} QEMU_PACKED;
+
+struct cdat_sslbe {
+    uint16_t port_x_id;
+    uint16_t port_y_id;
+    uint16_t latency_bandwidth;
+    uint16_t reserved;
+} QEMU_PACKED;
+
+struct cdat_sslbis_header {
+    struct cdat_sub_header header;
+    uint8_t data_type;
+    uint8_t reserved2[3];
+    uint64_t entry_base_unit;
+} QEMU_PACKED;
diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
new file mode 100644
index 0000000..ebbe488
--- /dev/null
+++ b/include/hw/cxl/cxl_compl.h
@@ -0,0 +1,289 @@
+/*
+ * CXL Compliance Mode Protocol
+ */
+struct cxl_compliance_mode_cap {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_cap_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+    uint64_t available_cap_bitmask;
+    uint64_t enabled_cap_bitmask;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_status {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_status_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint32_t cap_bitfield;
+    uint16_t cache_size;
+    uint8_t cache_size_units;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_halt {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_halt_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_multiple_write_streaming {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t protocol;
+    uint8_t virtual_addr;
+    uint8_t self_checking;
+    uint8_t verify_read_semantics;
+    uint8_t num_inc;
+    uint8_t num_sets;
+    uint8_t num_loops;
+    uint8_t reserved2;
+    uint64_t start_addr;
+    uint64_t write_addr;
+    uint64_t writeback_addr;
+    uint64_t byte_mask;
+    uint32_t addr_incr;
+    uint32_t set_offset;
+    uint32_t pattern_p;
+    uint32_t inc_pattern_b;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_multiple_write_streaming_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_producer_consumer {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t protocol;
+    uint8_t num_inc;
+    uint8_t num_sets;
+    uint8_t num_loops;
+    uint8_t write_semantics;
+    char reserved2[3];
+    uint64_t start_addr;
+    uint64_t byte_mask;
+    uint32_t addr_incr;
+    uint32_t set_offset;
+    uint32_t pattern;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_producer_consumer_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_bogus_writes {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t count;
+    uint8_t reserved2;
+    uint32_t pattern;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_bogus_writes_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_poison {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t protocol;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_poison_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_crc {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t num_bits_flip;
+    uint8_t num_flits_inj;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_crc_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_flow_control {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t inj_flow_control;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_flow_control_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_toggle_cache_flush {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t inj_flow_control;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_toggle_cache_flush_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_mac_delay {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t enable;
+    uint8_t mode;
+    uint8_t delay;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_mac_delay_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_insert_unexp_mac {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t opcode;
+    uint8_t mode;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_insert_unexp_mac_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_viral {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t protocol;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_viral_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t length;
+    uint8_t status;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_almp {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t opcode;
+    char reserved2[3];
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_inject_almp_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t reserved[6];
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_ignore_almp {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t opcode;
+    uint8_t reserved2[3];
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_ignore_almp_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t reserved[6];
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_ignore_bit_error {
+    DOEHeader header;
+    uint8_t req_code;
+    uint8_t version;
+    uint16_t reserved;
+    uint8_t opcode;
+} QEMU_PACKED;
+
+struct cxl_compliance_mode_ignore_bit_error_rsp {
+    DOEHeader header;
+    uint8_t rsp_code;
+    uint8_t version;
+    uint8_t reserved[6];
+} QEMU_PACKED;
diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
index 762feb5..23923df 100644
--- a/include/hw/cxl/cxl_component.h
+++ b/include/hw/cxl/cxl_component.h
@@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
 _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
                "No space for registers");
 
+/* 14.16.4 - Compliance Mode */
+#define CXL_COMP_MODE_CAP               0x0
+#define CXL_COMP_MODE_STATUS            0x1
+#define CXL_COMP_MODE_HALT              0x2
+#define CXL_COMP_MODE_MULT_WR_STREAM    0x3
+#define CXL_COMP_MODE_PRO_CON           0x4
+#define CXL_COMP_MODE_BOGUS             0x5
+#define CXL_COMP_MODE_INJ_POISON        0x6
+#define CXL_COMP_MODE_INJ_CRC           0x7
+#define CXL_COMP_MODE_INJ_FC            0x8
+#define CXL_COMP_MODE_TOGGLE_CACHE      0x9
+#define CXL_COMP_MODE_INJ_MAC           0xa
+#define CXL_COMP_MODE_INS_UNEXP_MAC     0xb
+#define CXL_COMP_MODE_INJ_VIRAL         0xc
+#define CXL_COMP_MODE_INJ_ALMP          0xd
+#define CXL_COMP_MODE_IGN_ALMP          0xe
+
 typedef struct component_registers {
     /*
      * Main memory region to be registered with QEMU core.
@@ -160,6 +177,10 @@ typedef struct component_registers {
     MemoryRegionOps *special_ops;
 } ComponentRegisters;
 
+typedef struct cdat_struct {
+    void *base;
+    uint32_t length;
+} CDATStruct;
 /*
  * A CXL component represents all entities in a CXL hierarchy. This includes,
  * host bridges, root ports, upstream/downstream switch ports, and devices
@@ -173,6 +194,104 @@ typedef struct cxl_component {
             struct PCIDevice *pdev;
         };
     };
+
+    /*
+     * SW write write mailbox and GO on last DW
+     * device sets READY of offset DW in DO types to copy
+     * to read mailbox register on subsequent cfg_read
+     * of read mailbox register and then on cfg_write to
+     * denote success read increments offset to next DW
+     */
+
+    union doe_request_u {
+        /* Compliance DOE Data Objects Type=0*/
+        struct cxl_compliance_mode_cap
+            mode_cap;
+        struct cxl_compliance_mode_status
+            mode_status;
+        struct cxl_compliance_mode_halt
+            mode_halt;
+        struct cxl_compliance_mode_multiple_write_streaming
+            multiple_write_streaming;
+        struct cxl_compliance_mode_producer_consumer
+            producer_consumer;
+        struct cxl_compliance_mode_inject_bogus_writes
+            inject_bogus_writes;
+        struct cxl_compliance_mode_inject_poison
+            inject_poison;
+        struct cxl_compliance_mode_inject_crc
+            inject_crc;
+        struct cxl_compliance_mode_inject_flow_control
+            inject_flow_control;
+        struct cxl_compliance_mode_toggle_cache_flush
+            toggle_cache_flush;
+        struct cxl_compliance_mode_inject_mac_delay
+            inject_mac_delay;
+        struct cxl_compliance_mode_insert_unexp_mac
+            insert_unexp_mac;
+        struct cxl_compliance_mode_inject_viral
+            inject_viral;
+        struct cxl_compliance_mode_inject_almp
+            inject_almp;
+        struct cxl_compliance_mode_ignore_almp
+            ignore_almp;
+        struct cxl_compliance_mode_ignore_bit_error
+            ignore_bit_error;
+        char data_byte[128];
+    } doe_request;
+
+    union doe_resp_u {
+        /* Compliance DOE Data Objects Type=0*/
+        struct cxl_compliance_mode_cap_rsp
+            cap_rsp;
+        struct cxl_compliance_mode_status_rsp
+            status_rsp;
+        struct cxl_compliance_mode_halt_rsp
+            halt_rsp;
+        struct cxl_compliance_mode_multiple_write_streaming_rsp
+            multiple_write_streaming_rsp;
+        struct cxl_compliance_mode_producer_consumer_rsp
+            producer_consumer_rsp;
+        struct cxl_compliance_mode_inject_bogus_writes_rsp
+            inject_bogus_writes_rsp;
+        struct cxl_compliance_mode_inject_poison_rsp
+            inject_poison_rsp;
+        struct cxl_compliance_mode_inject_crc_rsp
+            inject_crc_rsp;
+        struct cxl_compliance_mode_inject_flow_control_rsp
+            inject_flow_control_rsp;
+        struct cxl_compliance_mode_toggle_cache_flush_rsp
+            toggle_cache_flush_rsp;
+        struct cxl_compliance_mode_inject_mac_delay_rsp
+            inject_mac_delay_rsp;
+        struct cxl_compliance_mode_insert_unexp_mac_rsp
+            insert_unexp_mac_rsp;
+        struct cxl_compliance_mode_inject_viral
+            inject_viral_rsp;
+        struct cxl_compliance_mode_inject_almp_rsp
+            inject_almp_rsp;
+        struct cxl_compliance_mode_ignore_almp_rsp
+            ignore_almp_rsp;
+        struct cxl_compliance_mode_ignore_bit_error_rsp
+            ignore_bit_error_rsp;
+        char data_byte[520 * 4];
+        uint32_t data_dword[520];
+    } doe_resp;
+
+    /* Table entry types */
+    struct cdat_table_header cdat_header;
+    struct cdat_dsmas dsmas;
+    struct cdat_dslbis dslbis;
+    struct cdat_dsmscis dsmscis;
+    struct cdat_dsis dsis;
+    struct cdat_dsemts dsemts;
+    struct {
+        struct cdat_sslbis_header sslbis_h;
+        struct cdat_sslbe sslbe[3];
+    } sslbis;
+
+    CDATStruct *cdat_ent;
+    int cdat_ent_len;
 } CXLComponentState;
 
 void cxl_component_register_block_init(Object *obj,
@@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
 void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
                                 uint16_t type, uint8_t rev, uint8_t *body);
 
+void cxl_component_create_doe(CXLComponentState *cxl_cstate,
+                              uint16_t offset, unsigned vec);
+uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
+bool cxl_doe_compliance_rsp(DOECap *doe_cap);
+void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
+bool cxl_doe_cdat_rsp(DOECap *doe_cap);
+uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);
 #endif
diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
index c608ced..08bf646 100644
--- a/include/hw/cxl/cxl_device.h
+++ b/include/hw/cxl/cxl_device.h
@@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
     /* State */
     CXLComponentState cxl_cstate;
     CXLDeviceState cxl_dstate;
+
+    /* DOE */
+    PCIEDOE doe;
 } CXLType3Dev;
 
 #ifndef TYPE_CXL_TYPE3_DEV
diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
index 9ec28c9..5cab197 100644
--- a/include/hw/cxl/cxl_pci.h
+++ b/include/hw/cxl/cxl_pci.h
@@ -12,6 +12,8 @@
 
 #include "hw/pci/pci.h"
 #include "hw/pci/pcie.h"
+#include "hw/cxl/cxl_cdat.h"
+#include "hw/cxl/cxl_compl.h"
 
 #define CXL_VENDOR_ID 0x1e98
 
@@ -54,6 +56,8 @@ struct dvsec_header {
 _Static_assert(sizeof(struct dvsec_header) == 10,
                "dvsec header size incorrect");
 
+/* CXL 2.0 - 8.1.11 */
+
 /*
  * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
  * implement others.
-- 
1.8.3.1

Re: [RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Ben Widawsky]

A couple of high level comments below. Overall your approach was what I had
imagined originally. The approach Jonathan took is likely more versatile (but
harder to read, for sure).

I'm fine with either and I hope you two can come to an agreement on what the
best way forward is.

My ultimate goal was to be able to take a CDAT from a real device and load it as
a blob into the ct3d for regression testing. Not sure if that's actually
possible or not.

Thanks.
Ben

On 21-02-09 15:36:03, Chris Browy wrote:
> ---
>  hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
>  hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
>  include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
>  include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
>  include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
>  include/hw/cxl/cxl_device.h    |   3 +
>  include/hw/cxl/cxl_pci.h       |   4 +
>  7 files changed, 846 insertions(+)
>  create mode 100644 include/hw/cxl/cxl_cdat.h
>  create mode 100644 include/hw/cxl/cxl_compl.h
> 
> diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
> index e1bcee5..fc6c538 100644
> --- a/hw/cxl/cxl-component-utils.c
> +++ b/hw/cxl/cxl-component-utils.c
> @@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
>      range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
>      cxl->dvsec_offset += length;
>  }
> +
> +/* Return the sum of bytes */
> +static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
> +{
> +    cs->base = base;
> +    cs->length = len;
> +}
> +
> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
> +{
> +    uint8_t sum = 0;
> +    uint32_t len = 0;
> +    int i, j;
> +
> +    cxl_cstate->cdat_ent_len = 7;
> +    cxl_cstate->cdat_ent =
> +        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
> +
> +    cdat_ent_init(&cxl_cstate->cdat_ent[0],
> +                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[1],
> +                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[2],
> +                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[3],
> +                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[4],
> +                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[5],
> +                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[6],
> +                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
> +
> +    /* Set the DSMAS entry, ent = 1 */
> +    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
> +    cxl_cstate->dsmas.header.reserved = 0x0;
> +    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
> +    cxl_cstate->dsmas.DSMADhandle = 0x0;
> +    cxl_cstate->dsmas.flags = 0x0;
> +    cxl_cstate->dsmas.reserved2 = 0x0;
> +    cxl_cstate->dsmas.DPA_base = 0x0;
> +    cxl_cstate->dsmas.DPA_length = 0x40000;
> +
> +    /* Set the DSLBIS entry, ent = 2 */
> +    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
> +    cxl_cstate->dslbis.header.reserved = 0;
> +    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
> +    cxl_cstate->dslbis.handle = 0;
> +    cxl_cstate->dslbis.flags = 0;
> +    cxl_cstate->dslbis.data_type = 0;
> +    cxl_cstate->dslbis.reserved2 = 0;
> +    cxl_cstate->dslbis.entry_base_unit = 0;
> +    cxl_cstate->dslbis.entry[0] = 0;
> +    cxl_cstate->dslbis.entry[1] = 0;
> +    cxl_cstate->dslbis.entry[2] = 0;
> +    cxl_cstate->dslbis.reserved3 = 0;
> +
> +    /* Set the DSMSCIS entry, ent = 3 */
> +    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
> +    cxl_cstate->dsmscis.header.reserved = 0;
> +    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
> +    cxl_cstate->dsmscis.DSMASH_handle = 0;
> +    cxl_cstate->dsmscis.reserved2[0] = 0;
> +    cxl_cstate->dsmscis.reserved2[1] = 0;
> +    cxl_cstate->dsmscis.reserved2[2] = 0;
> +    cxl_cstate->dsmscis.memory_side_cache_size = 0;
> +    cxl_cstate->dsmscis.cache_attributes = 0;
> +
> +    /* Set the DSIS entry, ent = 4 */
> +    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
> +    cxl_cstate->dsis.header.reserved = 0;
> +    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
> +    cxl_cstate->dsis.flags = 0;
> +    cxl_cstate->dsis.handle = 0;
> +    cxl_cstate->dsis.reserved2 = 0;
> +
> +    /* Set the DSEMTS entry, ent = 5 */
> +    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
> +    cxl_cstate->dsemts.header.reserved = 0;
> +    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
> +    cxl_cstate->dsemts.DSMAS_handle = 0;
> +    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
> +    cxl_cstate->dsemts.reserved2 = 0;
> +    cxl_cstate->dsemts.DPA_offset = 0;
> +    cxl_cstate->dsemts.DPA_length = 0;
> +
> +    /* Set the SSLBIS entry, ent = 6 */
> +    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
> +    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
> +    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
> +    cxl_cstate->sslbis.sslbis_h.data_type = 0;
> +    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
> +    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
> +    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
> +    /* Set the SSLBE entry */
> +    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
> +    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
> +    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
> +    cxl_cstate->sslbis.sslbe[0].reserved = 0;
> +    /* Set the SSLBE entry */
> +    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
> +    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
> +    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
> +    cxl_cstate->sslbis.sslbe[1].reserved = 0;
> +    /* Set the SSLBE entry */
> +    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
> +    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
> +    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
> +    cxl_cstate->sslbis.sslbe[2].reserved = 0;
> +
> +    /* Set CDAT header, ent = 0 */
> +    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
> +    cxl_cstate->cdat_header.reserved[0] = 0;
> +    cxl_cstate->cdat_header.reserved[1] = 0;
> +    cxl_cstate->cdat_header.reserved[2] = 0;
> +    cxl_cstate->cdat_header.reserved[3] = 0;
> +    cxl_cstate->cdat_header.reserved[4] = 0;
> +    cxl_cstate->cdat_header.reserved[5] = 0;
> +    cxl_cstate->cdat_header.sequence = 0;
> +
> +    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
> +        /* Add length of each CDAT struct to total length */
> +        len = cxl_cstate->cdat_ent[i].length;
> +        cxl_cstate->cdat_header.length += len;
> +
> +        /* Calculate checksum of each CDAT struct */
> +        for (j = 0; j < len; j++) {
> +            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
> +        }
> +    }
> +    cxl_cstate->cdat_header.checksum = ~sum + 1;
> +}
> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
> index d091e64..86c762f 100644
> --- a/hw/mem/cxl_type3.c
> +++ b/hw/mem/cxl_type3.c
> @@ -13,6 +13,150 @@
>  #include "qemu/rcu.h"
>  #include "sysemu/hostmem.h"
>  #include "hw/cxl/cxl.h"
> +#include "hw/pci/msi.h"
> +#include "hw/pci/msix.h"
> +
> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap)
> +{
> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> +    uint32_t req;
> +    uint32_t byte_cnt = 0;
> +
> +    DOE_DBG(">> %s\n",  __func__);
> +
> +    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
> +        ->req_code;
> +    switch (req) {
> +    case CXL_COMP_MODE_CAP:
> +        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);
> +        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
> +        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
> +        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
> +        cxl_cstate->doe_resp.cap_rsp.header.length =
> +            dwsizeof(struct cxl_compliance_mode_cap_rsp);
> +        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
> +        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
> +        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
> +        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
> +        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
> +        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
> +        break;
> +    case CXL_COMP_MODE_STATUS:
> +        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
> +        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
> +        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
> +        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
> +        cxl_cstate->doe_resp.status_rsp.header.length =
> +            dwsizeof(struct cxl_compliance_mode_status_rsp);
> +        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
> +        cxl_cstate->doe_resp.status_rsp.version = 0x1;
> +        cxl_cstate->doe_resp.status_rsp.length = 0x14;
> +        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
> +        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
> +        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
> +        break;
> +    default:
> +        break;
> +    }
> +
> +    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
> +    DOE_DBG("<< %s\n",  __func__);
> +    return byte_cnt;
> +}
> +
> +
> +bool cxl_doe_compliance_rsp(DOECap *doe_cap)
> +{
> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> +    uint32_t byte_cnt;
> +    uint32_t dw_cnt;
> +
> +    DOE_DBG(">> %s\n",  __func__);
> +
> +    byte_cnt = cxl_doe_compliance_init(doe_cap);
> +    dw_cnt = byte_cnt / 4;
> +    memcpy(doe_cap->read_mbox,
> +           cxl_cstate->doe_resp.data_byte, byte_cnt);
> +
> +    doe_cap->read_mbox_len += dw_cnt;
> +
> +    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
> +            doe_cap->read_mbox_len,
> +            *(doe_cap->read_mbox + dw_cnt - 1));
> +
> +    DOE_DBG("<< %s\n",  __func__);
> +
> +    return DOE_SUCCESS;
> +}
> +
> +bool cxl_doe_cdat_rsp(DOECap *doe_cap)
> +{
> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> +    uint16_t ent;
> +    void *base;
> +    uint32_t len;
> +    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
> +
> +    ent = req->entry_handle;
> +    base = cxl_cstate->cdat_ent[ent].base;
> +    len = cxl_cstate->cdat_ent[ent].length;
> +
> +    struct cxl_cdat_rsp rsp = {
> +        .header = {
> +            .vendor_id = CXL_VENDOR_ID,
> +            .doe_type = CXL_DOE_TABLE_ACCESS,
> +            .reserved = 0x0,
> +            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
> +        },
> +        .req_code = CXL_DOE_TAB_RSP,
> +        .table_type = CXL_DOE_TAB_TYPE_CDAT,
> +        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
> +    };
> +
> +    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
> +    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
> +
> +    doe_cap->read_mbox_len += rsp.header.length;
> +    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
> +
> +    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
> +        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
> +    }
> +
> +    return DOE_SUCCESS;
> +}
> +
> +static uint32_t ct3d_config_read(PCIDevice *pci_dev,
> +                            uint32_t addr, int size)
> +{
> +    CXLType3Dev *ct3d = CT3(pci_dev);
> +    PCIEDOE *doe = &ct3d->doe;
> +    DOECap *doe_cap;
> +
> +    doe_cap = pcie_doe_covers_addr(doe, addr);
> +    if (doe_cap) {
> +        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
> +        return pcie_doe_read_config(doe_cap, addr, size);
> +    } else {
> +        return pci_default_read_config(pci_dev, addr, size);
> +    }
> +}
> +
> +static void ct3d_config_write(PCIDevice *pci_dev,
> +                            uint32_t addr, uint32_t val, int size)
> +{
> +    CXLType3Dev *ct3d = CT3(pci_dev);
> +    PCIEDOE *doe = &ct3d->doe;
> +    DOECap *doe_cap;
> +
> +    doe_cap = pcie_doe_covers_addr(doe, addr);
> +    if (doe_cap) {
> +        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
> +        pcie_doe_write_config(doe_cap, addr, val, size);
> +    } else {
> +        pci_default_write_config(pci_dev, addr, val, size);
> +    }
> +}
>  
>  static void build_dvsecs(CXLType3Dev *ct3d)
>  {
> @@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>      ComponentRegisters *regs = &cxl_cstate->crb;
>      MemoryRegion *mr = &regs->component_registers;
>      uint8_t *pci_conf = pci_dev->config;
> +    unsigned short msix_num = 2;
> +    int rc;
> +    int i;
>  
>      if (!ct3d->cxl_dstate.pmem) {
>          cxl_setup_memory(ct3d, errp);
> @@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>                       PCI_BASE_ADDRESS_SPACE_MEMORY |
>                           PCI_BASE_ADDRESS_MEM_TYPE_64,
>                       &ct3d->cxl_dstate.device_registers);
> +
> +    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
> +    for (i = 0; i < msix_num; i++) {
> +        msix_vector_use(pci_dev, i);
> +    }
> +
> +    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */
> +
> +    pcie_doe_init(pci_dev, &ct3d->doe);
> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);
> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
> +    if (rc) {
> +        error_setg(errp, "fail to add DOE cap");
> +        return;
> +    }
> +
> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
> +                               cxl_doe_compliance_rsp);
> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
> +                               cxl_doe_cdat_rsp);
> +
> +    cxl_doe_cdat_init(cxl_cstate);

So presumably you've looked at the way Jonathan does this. I like the idea of
being able to have a generic CDAT instantiation, but I haven't figured out if
it's realistically feasible. Please coordinate with him on this.

>  }
>  
>  static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
> @@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
>      DeviceClass *dc = DEVICE_CLASS(oc);
>      PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
>      MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
> +
> +    pc->config_write = ct3d_config_write;
> +    pc->config_read = ct3d_config_read;
>      CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
>  
>      pc->realize = ct3_realize;
> diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
> new file mode 100644
> index 0000000..fbbd494
> --- /dev/null
> +++ b/include/hw/cxl/cxl_cdat.h
> @@ -0,0 +1,120 @@
> +#include "hw/cxl/cxl_pci.h"
> +
> +
> +enum {
> +    CXL_DOE_COMPLIANCE             = 0,
> +    CXL_DOE_TABLE_ACCESS           = 2,
> +    CXL_DOE_MAX_PROTOCOL
> +};
> +
> +#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
> +#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
> +
> +/*
> + * DOE CDAT Table Protocol (CXL Spec)
> + */
> +#define CXL_DOE_TAB_REQ 0
> +#define CXL_DOE_TAB_RSP 0
> +#define CXL_DOE_TAB_TYPE_CDAT 0
> +#define CXL_DOE_TAB_ENT_MAX 0xFFFF
> +
> +/* Read Entry Request, 8.1.11.1 Table 134 */
> +struct cxl_cdat {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t table_type;
> +    uint16_t entry_handle;
> +} QEMU_PACKED;
> +
> +/* Read Entry Response, 8.1.11.1 Table 135 */
> +#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
> +
> +/*
> + * CDAT Table Structure (CDAT Spec)
> + */
> +#define CXL_CDAT_REV 1
> +
> +/* Data object header */
> +struct cdat_table_header {
> +    uint32_t length;    /* Length of table in bytes, including this header */
> +    uint8_t revision;   /* ACPI Specification minor version number */
> +    uint8_t checksum;   /* To make sum of entire table == 0 */
> +    uint8_t reserved[6];
> +    uint32_t sequence;  /* ASCII table signature */
> +} QEMU_PACKED;
> +
> +/* Values for subtable type in CDAT structures */
> +enum cdat_type {
> +    CDAT_TYPE_DSMAS = 0,
> +    CDAT_TYPE_DSLBIS = 1,
> +    CDAT_TYPE_DSMSCIS = 2,
> +    CDAT_TYPE_DSIS = 3,
> +    CDAT_TYPE_DSEMTS = 4,
> +    CDAT_TYPE_SSLBIS = 5,
> +    CDAT_TYPE_MAX
> +};
> +
> +struct cdat_sub_header {
> +    uint8_t type;
> +    uint8_t reserved;
> +    uint16_t length;
> +};
> +
> +/* CDAT Structure Subtables */
> +struct cdat_dsmas {
> +    struct cdat_sub_header header;
> +    uint8_t DSMADhandle;
> +    uint8_t flags;
> +    uint16_t reserved2;
> +    uint64_t DPA_base;
> +    uint64_t DPA_length;
> +} QEMU_PACKED;
> +
> +struct cdat_dslbis {
> +    struct cdat_sub_header header;
> +    uint8_t handle;
> +    uint8_t flags;
> +    uint8_t data_type;
> +    uint8_t reserved2;
> +    uint64_t entry_base_unit;
> +    uint16_t entry[3];
> +    uint16_t reserved3;
> +} QEMU_PACKED;
> +
> +struct cdat_dsmscis {
> +    struct cdat_sub_header header;
> +    uint8_t DSMASH_handle;
> +    uint8_t reserved2[3];
> +    uint64_t memory_side_cache_size;
> +    uint32_t cache_attributes;
> +} QEMU_PACKED;
> +
> +struct cdat_dsis {
> +    struct cdat_sub_header header;
> +    uint8_t flags;
> +    uint8_t handle;
> +    uint16_t reserved2;
> +} QEMU_PACKED;
> +
> +struct cdat_dsemts {
> +    struct cdat_sub_header header;
> +    uint8_t DSMAS_handle;
> +    uint8_t EFI_memory_type_attr;
> +    uint16_t reserved2;
> +    uint64_t DPA_offset;
> +    uint64_t DPA_length;
> +} QEMU_PACKED;
> +
> +struct cdat_sslbe {
> +    uint16_t port_x_id;
> +    uint16_t port_y_id;
> +    uint16_t latency_bandwidth;
> +    uint16_t reserved;
> +} QEMU_PACKED;
> +
> +struct cdat_sslbis_header {
> +    struct cdat_sub_header header;
> +    uint8_t data_type;
> +    uint8_t reserved2[3];
> +    uint64_t entry_base_unit;
> +} QEMU_PACKED;
> diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
> new file mode 100644
> index 0000000..ebbe488
> --- /dev/null
> +++ b/include/hw/cxl/cxl_compl.h
> @@ -0,0 +1,289 @@
> +/*
> + * CXL Compliance Mode Protocol
> + */
> +struct cxl_compliance_mode_cap {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_cap_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +    uint64_t available_cap_bitmask;
> +    uint64_t enabled_cap_bitmask;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_status {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_status_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint32_t cap_bitfield;
> +    uint16_t cache_size;
> +    uint8_t cache_size_units;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_halt {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_halt_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_multiple_write_streaming {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t protocol;
> +    uint8_t virtual_addr;
> +    uint8_t self_checking;
> +    uint8_t verify_read_semantics;
> +    uint8_t num_inc;
> +    uint8_t num_sets;
> +    uint8_t num_loops;
> +    uint8_t reserved2;
> +    uint64_t start_addr;
> +    uint64_t write_addr;
> +    uint64_t writeback_addr;
> +    uint64_t byte_mask;
> +    uint32_t addr_incr;
> +    uint32_t set_offset;
> +    uint32_t pattern_p;
> +    uint32_t inc_pattern_b;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_multiple_write_streaming_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_producer_consumer {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t protocol;
> +    uint8_t num_inc;
> +    uint8_t num_sets;
> +    uint8_t num_loops;
> +    uint8_t write_semantics;
> +    char reserved2[3];
> +    uint64_t start_addr;
> +    uint64_t byte_mask;
> +    uint32_t addr_incr;
> +    uint32_t set_offset;
> +    uint32_t pattern;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_producer_consumer_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_bogus_writes {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t count;
> +    uint8_t reserved2;
> +    uint32_t pattern;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_bogus_writes_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_poison {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t protocol;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_poison_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_crc {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t num_bits_flip;
> +    uint8_t num_flits_inj;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_crc_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_flow_control {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t inj_flow_control;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_flow_control_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_toggle_cache_flush {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t inj_flow_control;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_toggle_cache_flush_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_mac_delay {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t enable;
> +    uint8_t mode;
> +    uint8_t delay;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_mac_delay_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_insert_unexp_mac {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t opcode;
> +    uint8_t mode;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_insert_unexp_mac_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_viral {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t protocol;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_viral_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_almp {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t opcode;
> +    char reserved2[3];
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_inject_almp_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t reserved[6];
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_ignore_almp {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t opcode;
> +    uint8_t reserved2[3];
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_ignore_almp_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t reserved[6];
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_ignore_bit_error {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t opcode;
> +} QEMU_PACKED;
> +
> +struct cxl_compliance_mode_ignore_bit_error_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t reserved[6];
> +} QEMU_PACKED;

As this grows, I think it'd make sense to have a single header file with just
defines from the spec.

cxl_spec20.h or something like that. No need to change anything now, just a
thought.

> diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
> index 762feb5..23923df 100644
> --- a/include/hw/cxl/cxl_component.h
> +++ b/include/hw/cxl/cxl_component.h
> @@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
>  _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
>                 "No space for registers");
>  
> +/* 14.16.4 - Compliance Mode */
> +#define CXL_COMP_MODE_CAP               0x0
> +#define CXL_COMP_MODE_STATUS            0x1
> +#define CXL_COMP_MODE_HALT              0x2
> +#define CXL_COMP_MODE_MULT_WR_STREAM    0x3
> +#define CXL_COMP_MODE_PRO_CON           0x4
> +#define CXL_COMP_MODE_BOGUS             0x5
> +#define CXL_COMP_MODE_INJ_POISON        0x6
> +#define CXL_COMP_MODE_INJ_CRC           0x7
> +#define CXL_COMP_MODE_INJ_FC            0x8
> +#define CXL_COMP_MODE_TOGGLE_CACHE      0x9
> +#define CXL_COMP_MODE_INJ_MAC           0xa
> +#define CXL_COMP_MODE_INS_UNEXP_MAC     0xb
> +#define CXL_COMP_MODE_INJ_VIRAL         0xc
> +#define CXL_COMP_MODE_INJ_ALMP          0xd
> +#define CXL_COMP_MODE_IGN_ALMP          0xe
> +
>  typedef struct component_registers {
>      /*
>       * Main memory region to be registered with QEMU core.
> @@ -160,6 +177,10 @@ typedef struct component_registers {
>      MemoryRegionOps *special_ops;
>  } ComponentRegisters;
>  
> +typedef struct cdat_struct {
> +    void *base;
> +    uint32_t length;
> +} CDATStruct;
>  /*
>   * A CXL component represents all entities in a CXL hierarchy. This includes,
>   * host bridges, root ports, upstream/downstream switch ports, and devices
> @@ -173,6 +194,104 @@ typedef struct cxl_component {
>              struct PCIDevice *pdev;
>          };
>      };
> +
> +    /*
> +     * SW write write mailbox and GO on last DW
> +     * device sets READY of offset DW in DO types to copy
> +     * to read mailbox register on subsequent cfg_read
> +     * of read mailbox register and then on cfg_write to
> +     * denote success read increments offset to next DW
> +     */
> +
> +    union doe_request_u {
> +        /* Compliance DOE Data Objects Type=0*/
> +        struct cxl_compliance_mode_cap
> +            mode_cap;
> +        struct cxl_compliance_mode_status
> +            mode_status;
> +        struct cxl_compliance_mode_halt
> +            mode_halt;
> +        struct cxl_compliance_mode_multiple_write_streaming
> +            multiple_write_streaming;
> +        struct cxl_compliance_mode_producer_consumer
> +            producer_consumer;
> +        struct cxl_compliance_mode_inject_bogus_writes
> +            inject_bogus_writes;
> +        struct cxl_compliance_mode_inject_poison
> +            inject_poison;
> +        struct cxl_compliance_mode_inject_crc
> +            inject_crc;
> +        struct cxl_compliance_mode_inject_flow_control
> +            inject_flow_control;
> +        struct cxl_compliance_mode_toggle_cache_flush
> +            toggle_cache_flush;
> +        struct cxl_compliance_mode_inject_mac_delay
> +            inject_mac_delay;
> +        struct cxl_compliance_mode_insert_unexp_mac
> +            insert_unexp_mac;
> +        struct cxl_compliance_mode_inject_viral
> +            inject_viral;
> +        struct cxl_compliance_mode_inject_almp
> +            inject_almp;
> +        struct cxl_compliance_mode_ignore_almp
> +            ignore_almp;
> +        struct cxl_compliance_mode_ignore_bit_error
> +            ignore_bit_error;
> +        char data_byte[128];
> +    } doe_request;
> +
> +    union doe_resp_u {
> +        /* Compliance DOE Data Objects Type=0*/
> +        struct cxl_compliance_mode_cap_rsp
> +            cap_rsp;
> +        struct cxl_compliance_mode_status_rsp
> +            status_rsp;
> +        struct cxl_compliance_mode_halt_rsp
> +            halt_rsp;
> +        struct cxl_compliance_mode_multiple_write_streaming_rsp
> +            multiple_write_streaming_rsp;
> +        struct cxl_compliance_mode_producer_consumer_rsp
> +            producer_consumer_rsp;
> +        struct cxl_compliance_mode_inject_bogus_writes_rsp
> +            inject_bogus_writes_rsp;
> +        struct cxl_compliance_mode_inject_poison_rsp
> +            inject_poison_rsp;
> +        struct cxl_compliance_mode_inject_crc_rsp
> +            inject_crc_rsp;
> +        struct cxl_compliance_mode_inject_flow_control_rsp
> +            inject_flow_control_rsp;
> +        struct cxl_compliance_mode_toggle_cache_flush_rsp
> +            toggle_cache_flush_rsp;
> +        struct cxl_compliance_mode_inject_mac_delay_rsp
> +            inject_mac_delay_rsp;
> +        struct cxl_compliance_mode_insert_unexp_mac_rsp
> +            insert_unexp_mac_rsp;
> +        struct cxl_compliance_mode_inject_viral
> +            inject_viral_rsp;
> +        struct cxl_compliance_mode_inject_almp_rsp
> +            inject_almp_rsp;
> +        struct cxl_compliance_mode_ignore_almp_rsp
> +            ignore_almp_rsp;
> +        struct cxl_compliance_mode_ignore_bit_error_rsp
> +            ignore_bit_error_rsp;
> +        char data_byte[520 * 4];
> +        uint32_t data_dword[520];
> +    } doe_resp;
> +
> +    /* Table entry types */
> +    struct cdat_table_header cdat_header;
> +    struct cdat_dsmas dsmas;
> +    struct cdat_dslbis dslbis;
> +    struct cdat_dsmscis dsmscis;
> +    struct cdat_dsis dsis;
> +    struct cdat_dsemts dsemts;
> +    struct {
> +        struct cdat_sslbis_header sslbis_h;
> +        struct cdat_sslbe sslbe[3];
> +    } sslbis;
> +
> +    CDATStruct *cdat_ent;
> +    int cdat_ent_len;
>  } CXLComponentState;
>  
>  void cxl_component_register_block_init(Object *obj,
> @@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
>  void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
>                                  uint16_t type, uint8_t rev, uint8_t *body);
>  
> +void cxl_component_create_doe(CXLComponentState *cxl_cstate,
> +                              uint16_t offset, unsigned vec);
> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
> +bool cxl_doe_compliance_rsp(DOECap *doe_cap);
> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
> +bool cxl_doe_cdat_rsp(DOECap *doe_cap);
> +uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);
>  #endif
> diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
> index c608ced..08bf646 100644
> --- a/include/hw/cxl/cxl_device.h
> +++ b/include/hw/cxl/cxl_device.h
> @@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
>      /* State */
>      CXLComponentState cxl_cstate;
>      CXLDeviceState cxl_dstate;
> +
> +    /* DOE */
> +    PCIEDOE doe;
>  } CXLType3Dev;
>  
>  #ifndef TYPE_CXL_TYPE3_DEV
> diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
> index 9ec28c9..5cab197 100644
> --- a/include/hw/cxl/cxl_pci.h
> +++ b/include/hw/cxl/cxl_pci.h
> @@ -12,6 +12,8 @@
>  
>  #include "hw/pci/pci.h"
>  #include "hw/pci/pcie.h"
> +#include "hw/cxl/cxl_cdat.h"
> +#include "hw/cxl/cxl_compl.h"
>  
>  #define CXL_VENDOR_ID 0x1e98
>  
> @@ -54,6 +56,8 @@ struct dvsec_header {
>  _Static_assert(sizeof(struct dvsec_header) == 10,
>                 "dvsec header size incorrect");
>  
> +/* CXL 2.0 - 8.1.11 */
> +
>  /*
>   * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
>   * implement others.

Re: [RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Chris Browy]

> On Feb 9, 2021, at 4:53 PM, Ben Widawsky <ben@bwidawsk.net> wrote:
> 
> A couple of high level comments below. Overall your approach was what I had
> imagined originally. The approach Jonathan took is likely more versatile (but
> harder to read, for sure).
> 
> I'm fine with either and I hope you two can come to an agreement on what the
> best way forward is.
> 
> My ultimate goal was to be able to take a CDAT from a real device and load it as
> a blob into the ct3d for regression testing. Not sure if that's actually
> possible or not.

I’d think so.  

For CDAT/DOE method, you could setup CDAT as non-ACPI tables but compile 
with ACPI iASL?  UEFI owns the ACPI and CDAT specs and all the info is public.

For example using generic datatypes one can describe CDAT structure types and
create an arbitrary CDAT table with any mix of struct types and describe one or more proximity
domains and their memory attributes.  The ct3d device can read the “blob” or .aml and setup 
entry indexing as Jonathan mentioned previously.  For example user could create a
CDAT table and compile using iasl -G <file.dat> into a file.aml and disassemble back 
Into a file.dsl.

Here is example of CDAT header and DSMAS (with ACPI standard header as well):

Signature : "CDAT"
Table Length : 00000000
Revision : 01
Checksum : 00
Oem ID : "TEST"
Oem Table ID : "QEMU "
Oem Revision : 00000001
Asl Compiler ID : "INTL"
Asl Compiler Revision : 00000001

Label : CDATST
Label : CDAT_HDR
UINT32 : $CDATEND - $CDATST
UINT8  : 01                     // Revision             1
UINT8  : 00                     // Checksum             1
UINT24 : 000000                 // Reserved             6
UINT32 : 00000000               // Sequence             4

Label : DSMAS                   // Field                Byte Length
UINT8  : 00                     // Type                 1
UINT8  : 00                     // Reserved             1
UINT16 : 0018                   // Length               2
UINT8  : 00                     // DSMADHandle          1
UINT8  : 00                     // Flags                1
UINT16 : 0000                   // Reserved             2
UINT64 : 0000000000000000       // DPA Base             8
UINT64 : 0000000000000000       // DPA Length           8


For Device Option ROM method for CDAT, we could add a option rom to ct3d so UEFI could 
access CDAT through a  EFI_ADAPTER_INFORMATION_PROTOCOL (CDAT type) entry.


> 
> Thanks.
> Ben
> 
> On 21-02-09 15:36:03, Chris Browy wrote:
>> ---
>> hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
>> hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
>> include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
>> include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
>> include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
>> include/hw/cxl/cxl_device.h    |   3 +
>> include/hw/cxl/cxl_pci.h       |   4 +
>> 7 files changed, 846 insertions(+)
>> create mode 100644 include/hw/cxl/cxl_cdat.h
>> create mode 100644 include/hw/cxl/cxl_compl.h
>> 
>> diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
>> index e1bcee5..fc6c538 100644
>> --- a/hw/cxl/cxl-component-utils.c
>> +++ b/hw/cxl/cxl-component-utils.c
>> @@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
>>     range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
>>     cxl->dvsec_offset += length;
>> }
>> +
>> +/* Return the sum of bytes */
>> +static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
>> +{
>> +    cs->base = base;
>> +    cs->length = len;
>> +}
>> +
>> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
>> +{
>> +    uint8_t sum = 0;
>> +    uint32_t len = 0;
>> +    int i, j;
>> +
>> +    cxl_cstate->cdat_ent_len = 7;
>> +    cxl_cstate->cdat_ent =
>> +        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
>> +
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[0],
>> +                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[1],
>> +                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[2],
>> +                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[3],
>> +                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[4],
>> +                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[5],
>> +                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[6],
>> +                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
>> +
>> +    /* Set the DSMAS entry, ent = 1 */
>> +    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
>> +    cxl_cstate->dsmas.header.reserved = 0x0;
>> +    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
>> +    cxl_cstate->dsmas.DSMADhandle = 0x0;
>> +    cxl_cstate->dsmas.flags = 0x0;
>> +    cxl_cstate->dsmas.reserved2 = 0x0;
>> +    cxl_cstate->dsmas.DPA_base = 0x0;
>> +    cxl_cstate->dsmas.DPA_length = 0x40000;
>> +
>> +    /* Set the DSLBIS entry, ent = 2 */
>> +    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
>> +    cxl_cstate->dslbis.header.reserved = 0;
>> +    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
>> +    cxl_cstate->dslbis.handle = 0;
>> +    cxl_cstate->dslbis.flags = 0;
>> +    cxl_cstate->dslbis.data_type = 0;
>> +    cxl_cstate->dslbis.reserved2 = 0;
>> +    cxl_cstate->dslbis.entry_base_unit = 0;
>> +    cxl_cstate->dslbis.entry[0] = 0;
>> +    cxl_cstate->dslbis.entry[1] = 0;
>> +    cxl_cstate->dslbis.entry[2] = 0;
>> +    cxl_cstate->dslbis.reserved3 = 0;
>> +
>> +    /* Set the DSMSCIS entry, ent = 3 */
>> +    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
>> +    cxl_cstate->dsmscis.header.reserved = 0;
>> +    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
>> +    cxl_cstate->dsmscis.DSMASH_handle = 0;
>> +    cxl_cstate->dsmscis.reserved2[0] = 0;
>> +    cxl_cstate->dsmscis.reserved2[1] = 0;
>> +    cxl_cstate->dsmscis.reserved2[2] = 0;
>> +    cxl_cstate->dsmscis.memory_side_cache_size = 0;
>> +    cxl_cstate->dsmscis.cache_attributes = 0;
>> +
>> +    /* Set the DSIS entry, ent = 4 */
>> +    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
>> +    cxl_cstate->dsis.header.reserved = 0;
>> +    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
>> +    cxl_cstate->dsis.flags = 0;
>> +    cxl_cstate->dsis.handle = 0;
>> +    cxl_cstate->dsis.reserved2 = 0;
>> +
>> +    /* Set the DSEMTS entry, ent = 5 */
>> +    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
>> +    cxl_cstate->dsemts.header.reserved = 0;
>> +    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
>> +    cxl_cstate->dsemts.DSMAS_handle = 0;
>> +    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
>> +    cxl_cstate->dsemts.reserved2 = 0;
>> +    cxl_cstate->dsemts.DPA_offset = 0;
>> +    cxl_cstate->dsemts.DPA_length = 0;
>> +
>> +    /* Set the SSLBIS entry, ent = 6 */
>> +    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
>> +    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
>> +    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
>> +    cxl_cstate->sslbis.sslbis_h.data_type = 0;
>> +    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
>> +    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
>> +    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
>> +    /* Set the SSLBE entry */
>> +    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
>> +    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
>> +    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
>> +    cxl_cstate->sslbis.sslbe[0].reserved = 0;
>> +    /* Set the SSLBE entry */
>> +    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
>> +    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
>> +    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
>> +    cxl_cstate->sslbis.sslbe[1].reserved = 0;
>> +    /* Set the SSLBE entry */
>> +    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
>> +    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
>> +    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
>> +    cxl_cstate->sslbis.sslbe[2].reserved = 0;
>> +
>> +    /* Set CDAT header, ent = 0 */
>> +    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
>> +    cxl_cstate->cdat_header.reserved[0] = 0;
>> +    cxl_cstate->cdat_header.reserved[1] = 0;
>> +    cxl_cstate->cdat_header.reserved[2] = 0;
>> +    cxl_cstate->cdat_header.reserved[3] = 0;
>> +    cxl_cstate->cdat_header.reserved[4] = 0;
>> +    cxl_cstate->cdat_header.reserved[5] = 0;
>> +    cxl_cstate->cdat_header.sequence = 0;
>> +
>> +    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
>> +        /* Add length of each CDAT struct to total length */
>> +        len = cxl_cstate->cdat_ent[i].length;
>> +        cxl_cstate->cdat_header.length += len;
>> +
>> +        /* Calculate checksum of each CDAT struct */
>> +        for (j = 0; j < len; j++) {
>> +            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
>> +        }
>> +    }
>> +    cxl_cstate->cdat_header.checksum = ~sum + 1;
>> +}
>> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
>> index d091e64..86c762f 100644
>> --- a/hw/mem/cxl_type3.c
>> +++ b/hw/mem/cxl_type3.c
>> @@ -13,6 +13,150 @@
>> #include "qemu/rcu.h"
>> #include "sysemu/hostmem.h"
>> #include "hw/cxl/cxl.h"
>> +#include "hw/pci/msi.h"
>> +#include "hw/pci/msix.h"
>> +
>> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap)
>> +{
>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>> +    uint32_t req;
>> +    uint32_t byte_cnt = 0;
>> +
>> +    DOE_DBG(">> %s\n",  __func__);
>> +
>> +    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
>> +        ->req_code;
>> +    switch (req) {
>> +    case CXL_COMP_MODE_CAP:
>> +        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);
>> +        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
>> +        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
>> +        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
>> +        cxl_cstate->doe_resp.cap_rsp.header.length =
>> +            dwsizeof(struct cxl_compliance_mode_cap_rsp);
>> +        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
>> +        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
>> +        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
>> +        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
>> +        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
>> +        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
>> +        break;
>> +    case CXL_COMP_MODE_STATUS:
>> +        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
>> +        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
>> +        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
>> +        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
>> +        cxl_cstate->doe_resp.status_rsp.header.length =
>> +            dwsizeof(struct cxl_compliance_mode_status_rsp);
>> +        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
>> +        cxl_cstate->doe_resp.status_rsp.version = 0x1;
>> +        cxl_cstate->doe_resp.status_rsp.length = 0x14;
>> +        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
>> +        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
>> +        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
>> +        break;
>> +    default:
>> +        break;
>> +    }
>> +
>> +    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
>> +    DOE_DBG("<< %s\n",  __func__);
>> +    return byte_cnt;
>> +}
>> +
>> +
>> +bool cxl_doe_compliance_rsp(DOECap *doe_cap)
>> +{
>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>> +    uint32_t byte_cnt;
>> +    uint32_t dw_cnt;
>> +
>> +    DOE_DBG(">> %s\n",  __func__);
>> +
>> +    byte_cnt = cxl_doe_compliance_init(doe_cap);
>> +    dw_cnt = byte_cnt / 4;
>> +    memcpy(doe_cap->read_mbox,
>> +           cxl_cstate->doe_resp.data_byte, byte_cnt);
>> +
>> +    doe_cap->read_mbox_len += dw_cnt;
>> +
>> +    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
>> +            doe_cap->read_mbox_len,
>> +            *(doe_cap->read_mbox + dw_cnt - 1));
>> +
>> +    DOE_DBG("<< %s\n",  __func__);
>> +
>> +    return DOE_SUCCESS;
>> +}
>> +
>> +bool cxl_doe_cdat_rsp(DOECap *doe_cap)
>> +{
>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>> +    uint16_t ent;
>> +    void *base;
>> +    uint32_t len;
>> +    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
>> +
>> +    ent = req->entry_handle;
>> +    base = cxl_cstate->cdat_ent[ent].base;
>> +    len = cxl_cstate->cdat_ent[ent].length;
>> +
>> +    struct cxl_cdat_rsp rsp = {
>> +        .header = {
>> +            .vendor_id = CXL_VENDOR_ID,
>> +            .doe_type = CXL_DOE_TABLE_ACCESS,
>> +            .reserved = 0x0,
>> +            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
>> +        },
>> +        .req_code = CXL_DOE_TAB_RSP,
>> +        .table_type = CXL_DOE_TAB_TYPE_CDAT,
>> +        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
>> +    };
>> +
>> +    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
>> +    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
>> +
>> +    doe_cap->read_mbox_len += rsp.header.length;
>> +    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
>> +
>> +    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
>> +        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
>> +    }
>> +
>> +    return DOE_SUCCESS;
>> +}
>> +
>> +static uint32_t ct3d_config_read(PCIDevice *pci_dev,
>> +                            uint32_t addr, int size)
>> +{
>> +    CXLType3Dev *ct3d = CT3(pci_dev);
>> +    PCIEDOE *doe = &ct3d->doe;
>> +    DOECap *doe_cap;
>> +
>> +    doe_cap = pcie_doe_covers_addr(doe, addr);
>> +    if (doe_cap) {
>> +        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
>> +        return pcie_doe_read_config(doe_cap, addr, size);
>> +    } else {
>> +        return pci_default_read_config(pci_dev, addr, size);
>> +    }
>> +}
>> +
>> +static void ct3d_config_write(PCIDevice *pci_dev,
>> +                            uint32_t addr, uint32_t val, int size)
>> +{
>> +    CXLType3Dev *ct3d = CT3(pci_dev);
>> +    PCIEDOE *doe = &ct3d->doe;
>> +    DOECap *doe_cap;
>> +
>> +    doe_cap = pcie_doe_covers_addr(doe, addr);
>> +    if (doe_cap) {
>> +        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
>> +        pcie_doe_write_config(doe_cap, addr, val, size);
>> +    } else {
>> +        pci_default_write_config(pci_dev, addr, val, size);
>> +    }
>> +}
>> 
>> static void build_dvsecs(CXLType3Dev *ct3d)
>> {
>> @@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>>     ComponentRegisters *regs = &cxl_cstate->crb;
>>     MemoryRegion *mr = &regs->component_registers;
>>     uint8_t *pci_conf = pci_dev->config;
>> +    unsigned short msix_num = 2;
>> +    int rc;
>> +    int i;
>> 
>>     if (!ct3d->cxl_dstate.pmem) {
>>         cxl_setup_memory(ct3d, errp);
>> @@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>>                      PCI_BASE_ADDRESS_SPACE_MEMORY |
>>                          PCI_BASE_ADDRESS_MEM_TYPE_64,
>>                      &ct3d->cxl_dstate.device_registers);
>> +
>> +    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
>> +    for (i = 0; i < msix_num; i++) {
>> +        msix_vector_use(pci_dev, i);
>> +    }
>> +
>> +    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */
>> +
>> +    pcie_doe_init(pci_dev, &ct3d->doe);
>> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);
>> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
>> +    if (rc) {
>> +        error_setg(errp, "fail to add DOE cap");
>> +        return;
>> +    }
>> +
>> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
>> +                               cxl_doe_compliance_rsp);
>> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
>> +                               cxl_doe_cdat_rsp);
>> +
>> +    cxl_doe_cdat_init(cxl_cstate);
> 
> So presumably you've looked at the way Jonathan does this. I like the idea of
> being able to have a generic CDAT instantiation, but I haven't figured out if
> it's realistically feasible. Please coordinate with him on this.
> 
>> }
>> 
>> static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
>> @@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
>>     DeviceClass *dc = DEVICE_CLASS(oc);
>>     PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
>>     MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
>> +
>> +    pc->config_write = ct3d_config_write;
>> +    pc->config_read = ct3d_config_read;
>>     CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
>> 
>>     pc->realize = ct3_realize;
>> diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
>> new file mode 100644
>> index 0000000..fbbd494
>> --- /dev/null
>> +++ b/include/hw/cxl/cxl_cdat.h
>> @@ -0,0 +1,120 @@
>> +#include "hw/cxl/cxl_pci.h"
>> +
>> +
>> +enum {
>> +    CXL_DOE_COMPLIANCE             = 0,
>> +    CXL_DOE_TABLE_ACCESS           = 2,
>> +    CXL_DOE_MAX_PROTOCOL
>> +};
>> +
>> +#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
>> +#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
>> +
>> +/*
>> + * DOE CDAT Table Protocol (CXL Spec)
>> + */
>> +#define CXL_DOE_TAB_REQ 0
>> +#define CXL_DOE_TAB_RSP 0
>> +#define CXL_DOE_TAB_TYPE_CDAT 0
>> +#define CXL_DOE_TAB_ENT_MAX 0xFFFF
>> +
>> +/* Read Entry Request, 8.1.11.1 Table 134 */
>> +struct cxl_cdat {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t table_type;
>> +    uint16_t entry_handle;
>> +} QEMU_PACKED;
>> +
>> +/* Read Entry Response, 8.1.11.1 Table 135 */
>> +#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
>> +
>> +/*
>> + * CDAT Table Structure (CDAT Spec)
>> + */
>> +#define CXL_CDAT_REV 1
>> +
>> +/* Data object header */
>> +struct cdat_table_header {
>> +    uint32_t length;    /* Length of table in bytes, including this header */
>> +    uint8_t revision;   /* ACPI Specification minor version number */
>> +    uint8_t checksum;   /* To make sum of entire table == 0 */
>> +    uint8_t reserved[6];
>> +    uint32_t sequence;  /* ASCII table signature */
>> +} QEMU_PACKED;
>> +
>> +/* Values for subtable type in CDAT structures */
>> +enum cdat_type {
>> +    CDAT_TYPE_DSMAS = 0,
>> +    CDAT_TYPE_DSLBIS = 1,
>> +    CDAT_TYPE_DSMSCIS = 2,
>> +    CDAT_TYPE_DSIS = 3,
>> +    CDAT_TYPE_DSEMTS = 4,
>> +    CDAT_TYPE_SSLBIS = 5,
>> +    CDAT_TYPE_MAX
>> +};
>> +
>> +struct cdat_sub_header {
>> +    uint8_t type;
>> +    uint8_t reserved;
>> +    uint16_t length;
>> +};
>> +
>> +/* CDAT Structure Subtables */
>> +struct cdat_dsmas {
>> +    struct cdat_sub_header header;
>> +    uint8_t DSMADhandle;
>> +    uint8_t flags;
>> +    uint16_t reserved2;
>> +    uint64_t DPA_base;
>> +    uint64_t DPA_length;
>> +} QEMU_PACKED;
>> +
>> +struct cdat_dslbis {
>> +    struct cdat_sub_header header;
>> +    uint8_t handle;
>> +    uint8_t flags;
>> +    uint8_t data_type;
>> +    uint8_t reserved2;
>> +    uint64_t entry_base_unit;
>> +    uint16_t entry[3];
>> +    uint16_t reserved3;
>> +} QEMU_PACKED;
>> +
>> +struct cdat_dsmscis {
>> +    struct cdat_sub_header header;
>> +    uint8_t DSMASH_handle;
>> +    uint8_t reserved2[3];
>> +    uint64_t memory_side_cache_size;
>> +    uint32_t cache_attributes;
>> +} QEMU_PACKED;
>> +
>> +struct cdat_dsis {
>> +    struct cdat_sub_header header;
>> +    uint8_t flags;
>> +    uint8_t handle;
>> +    uint16_t reserved2;
>> +} QEMU_PACKED;
>> +
>> +struct cdat_dsemts {
>> +    struct cdat_sub_header header;
>> +    uint8_t DSMAS_handle;
>> +    uint8_t EFI_memory_type_attr;
>> +    uint16_t reserved2;
>> +    uint64_t DPA_offset;
>> +    uint64_t DPA_length;
>> +} QEMU_PACKED;
>> +
>> +struct cdat_sslbe {
>> +    uint16_t port_x_id;
>> +    uint16_t port_y_id;
>> +    uint16_t latency_bandwidth;
>> +    uint16_t reserved;
>> +} QEMU_PACKED;
>> +
>> +struct cdat_sslbis_header {
>> +    struct cdat_sub_header header;
>> +    uint8_t data_type;
>> +    uint8_t reserved2[3];
>> +    uint64_t entry_base_unit;
>> +} QEMU_PACKED;
>> diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
>> new file mode 100644
>> index 0000000..ebbe488
>> --- /dev/null
>> +++ b/include/hw/cxl/cxl_compl.h
>> @@ -0,0 +1,289 @@
>> +/*
>> + * CXL Compliance Mode Protocol
>> + */
>> +struct cxl_compliance_mode_cap {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_cap_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +    uint64_t available_cap_bitmask;
>> +    uint64_t enabled_cap_bitmask;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_status {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_status_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint32_t cap_bitfield;
>> +    uint16_t cache_size;
>> +    uint8_t cache_size_units;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_halt {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_halt_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_multiple_write_streaming {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t protocol;
>> +    uint8_t virtual_addr;
>> +    uint8_t self_checking;
>> +    uint8_t verify_read_semantics;
>> +    uint8_t num_inc;
>> +    uint8_t num_sets;
>> +    uint8_t num_loops;
>> +    uint8_t reserved2;
>> +    uint64_t start_addr;
>> +    uint64_t write_addr;
>> +    uint64_t writeback_addr;
>> +    uint64_t byte_mask;
>> +    uint32_t addr_incr;
>> +    uint32_t set_offset;
>> +    uint32_t pattern_p;
>> +    uint32_t inc_pattern_b;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_multiple_write_streaming_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_producer_consumer {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t protocol;
>> +    uint8_t num_inc;
>> +    uint8_t num_sets;
>> +    uint8_t num_loops;
>> +    uint8_t write_semantics;
>> +    char reserved2[3];
>> +    uint64_t start_addr;
>> +    uint64_t byte_mask;
>> +    uint32_t addr_incr;
>> +    uint32_t set_offset;
>> +    uint32_t pattern;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_producer_consumer_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_bogus_writes {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t count;
>> +    uint8_t reserved2;
>> +    uint32_t pattern;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_bogus_writes_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_poison {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t protocol;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_poison_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_crc {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t num_bits_flip;
>> +    uint8_t num_flits_inj;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_crc_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_flow_control {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t inj_flow_control;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_flow_control_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_toggle_cache_flush {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t inj_flow_control;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_toggle_cache_flush_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_mac_delay {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t enable;
>> +    uint8_t mode;
>> +    uint8_t delay;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_mac_delay_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_insert_unexp_mac {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t opcode;
>> +    uint8_t mode;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_insert_unexp_mac_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_viral {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t protocol;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_viral_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_almp {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t opcode;
>> +    char reserved2[3];
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_inject_almp_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t reserved[6];
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_ignore_almp {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t opcode;
>> +    uint8_t reserved2[3];
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_ignore_almp_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t reserved[6];
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_ignore_bit_error {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t opcode;
>> +} QEMU_PACKED;
>> +
>> +struct cxl_compliance_mode_ignore_bit_error_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t reserved[6];
>> +} QEMU_PACKED;
> 
> As this grows, I think it'd make sense to have a single header file with just
> defines from the spec.
> 
> cxl_spec20.h or something like that. No need to change anything now, just a
> thought.
> 
>> diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
>> index 762feb5..23923df 100644
>> --- a/include/hw/cxl/cxl_component.h
>> +++ b/include/hw/cxl/cxl_component.h
>> @@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
>> _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
>>                "No space for registers");
>> 
>> +/* 14.16.4 - Compliance Mode */
>> +#define CXL_COMP_MODE_CAP               0x0
>> +#define CXL_COMP_MODE_STATUS            0x1
>> +#define CXL_COMP_MODE_HALT              0x2
>> +#define CXL_COMP_MODE_MULT_WR_STREAM    0x3
>> +#define CXL_COMP_MODE_PRO_CON           0x4
>> +#define CXL_COMP_MODE_BOGUS             0x5
>> +#define CXL_COMP_MODE_INJ_POISON        0x6
>> +#define CXL_COMP_MODE_INJ_CRC           0x7
>> +#define CXL_COMP_MODE_INJ_FC            0x8
>> +#define CXL_COMP_MODE_TOGGLE_CACHE      0x9
>> +#define CXL_COMP_MODE_INJ_MAC           0xa
>> +#define CXL_COMP_MODE_INS_UNEXP_MAC     0xb
>> +#define CXL_COMP_MODE_INJ_VIRAL         0xc
>> +#define CXL_COMP_MODE_INJ_ALMP          0xd
>> +#define CXL_COMP_MODE_IGN_ALMP          0xe
>> +
>> typedef struct component_registers {
>>     /*
>>      * Main memory region to be registered with QEMU core.
>> @@ -160,6 +177,10 @@ typedef struct component_registers {
>>     MemoryRegionOps *special_ops;
>> } ComponentRegisters;
>> 
>> +typedef struct cdat_struct {
>> +    void *base;
>> +    uint32_t length;
>> +} CDATStruct;
>> /*
>>  * A CXL component represents all entities in a CXL hierarchy. This includes,
>>  * host bridges, root ports, upstream/downstream switch ports, and devices
>> @@ -173,6 +194,104 @@ typedef struct cxl_component {
>>             struct PCIDevice *pdev;
>>         };
>>     };
>> +
>> +    /*
>> +     * SW write write mailbox and GO on last DW
>> +     * device sets READY of offset DW in DO types to copy
>> +     * to read mailbox register on subsequent cfg_read
>> +     * of read mailbox register and then on cfg_write to
>> +     * denote success read increments offset to next DW
>> +     */
>> +
>> +    union doe_request_u {
>> +        /* Compliance DOE Data Objects Type=0*/
>> +        struct cxl_compliance_mode_cap
>> +            mode_cap;
>> +        struct cxl_compliance_mode_status
>> +            mode_status;
>> +        struct cxl_compliance_mode_halt
>> +            mode_halt;
>> +        struct cxl_compliance_mode_multiple_write_streaming
>> +            multiple_write_streaming;
>> +        struct cxl_compliance_mode_producer_consumer
>> +            producer_consumer;
>> +        struct cxl_compliance_mode_inject_bogus_writes
>> +            inject_bogus_writes;
>> +        struct cxl_compliance_mode_inject_poison
>> +            inject_poison;
>> +        struct cxl_compliance_mode_inject_crc
>> +            inject_crc;
>> +        struct cxl_compliance_mode_inject_flow_control
>> +            inject_flow_control;
>> +        struct cxl_compliance_mode_toggle_cache_flush
>> +            toggle_cache_flush;
>> +        struct cxl_compliance_mode_inject_mac_delay
>> +            inject_mac_delay;
>> +        struct cxl_compliance_mode_insert_unexp_mac
>> +            insert_unexp_mac;
>> +        struct cxl_compliance_mode_inject_viral
>> +            inject_viral;
>> +        struct cxl_compliance_mode_inject_almp
>> +            inject_almp;
>> +        struct cxl_compliance_mode_ignore_almp
>> +            ignore_almp;
>> +        struct cxl_compliance_mode_ignore_bit_error
>> +            ignore_bit_error;
>> +        char data_byte[128];
>> +    } doe_request;
>> +
>> +    union doe_resp_u {
>> +        /* Compliance DOE Data Objects Type=0*/
>> +        struct cxl_compliance_mode_cap_rsp
>> +            cap_rsp;
>> +        struct cxl_compliance_mode_status_rsp
>> +            status_rsp;
>> +        struct cxl_compliance_mode_halt_rsp
>> +            halt_rsp;
>> +        struct cxl_compliance_mode_multiple_write_streaming_rsp
>> +            multiple_write_streaming_rsp;
>> +        struct cxl_compliance_mode_producer_consumer_rsp
>> +            producer_consumer_rsp;
>> +        struct cxl_compliance_mode_inject_bogus_writes_rsp
>> +            inject_bogus_writes_rsp;
>> +        struct cxl_compliance_mode_inject_poison_rsp
>> +            inject_poison_rsp;
>> +        struct cxl_compliance_mode_inject_crc_rsp
>> +            inject_crc_rsp;
>> +        struct cxl_compliance_mode_inject_flow_control_rsp
>> +            inject_flow_control_rsp;
>> +        struct cxl_compliance_mode_toggle_cache_flush_rsp
>> +            toggle_cache_flush_rsp;
>> +        struct cxl_compliance_mode_inject_mac_delay_rsp
>> +            inject_mac_delay_rsp;
>> +        struct cxl_compliance_mode_insert_unexp_mac_rsp
>> +            insert_unexp_mac_rsp;
>> +        struct cxl_compliance_mode_inject_viral
>> +            inject_viral_rsp;
>> +        struct cxl_compliance_mode_inject_almp_rsp
>> +            inject_almp_rsp;
>> +        struct cxl_compliance_mode_ignore_almp_rsp
>> +            ignore_almp_rsp;
>> +        struct cxl_compliance_mode_ignore_bit_error_rsp
>> +            ignore_bit_error_rsp;
>> +        char data_byte[520 * 4];
>> +        uint32_t data_dword[520];
>> +    } doe_resp;
>> +
>> +    /* Table entry types */
>> +    struct cdat_table_header cdat_header;
>> +    struct cdat_dsmas dsmas;
>> +    struct cdat_dslbis dslbis;
>> +    struct cdat_dsmscis dsmscis;
>> +    struct cdat_dsis dsis;
>> +    struct cdat_dsemts dsemts;
>> +    struct {
>> +        struct cdat_sslbis_header sslbis_h;
>> +        struct cdat_sslbe sslbe[3];
>> +    } sslbis;
>> +
>> +    CDATStruct *cdat_ent;
>> +    int cdat_ent_len;
>> } CXLComponentState;
>> 
>> void cxl_component_register_block_init(Object *obj,
>> @@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
>> void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
>>                                 uint16_t type, uint8_t rev, uint8_t *body);
>> 
>> +void cxl_component_create_doe(CXLComponentState *cxl_cstate,
>> +                              uint16_t offset, unsigned vec);
>> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
>> +bool cxl_doe_compliance_rsp(DOECap *doe_cap);
>> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
>> +bool cxl_doe_cdat_rsp(DOECap *doe_cap);
>> +uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);
>> #endif
>> diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
>> index c608ced..08bf646 100644
>> --- a/include/hw/cxl/cxl_device.h
>> +++ b/include/hw/cxl/cxl_device.h
>> @@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
>>     /* State */
>>     CXLComponentState cxl_cstate;
>>     CXLDeviceState cxl_dstate;
>> +
>> +    /* DOE */
>> +    PCIEDOE doe;
>> } CXLType3Dev;
>> 
>> #ifndef TYPE_CXL_TYPE3_DEV
>> diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
>> index 9ec28c9..5cab197 100644
>> --- a/include/hw/cxl/cxl_pci.h
>> +++ b/include/hw/cxl/cxl_pci.h
>> @@ -12,6 +12,8 @@
>> 
>> #include "hw/pci/pci.h"
>> #include "hw/pci/pcie.h"
>> +#include "hw/cxl/cxl_cdat.h"
>> +#include "hw/cxl/cxl_compl.h"
>> 
>> #define CXL_VENDOR_ID 0x1e98
>> 
>> @@ -54,6 +56,8 @@ struct dvsec_header {
>> _Static_assert(sizeof(struct dvsec_header) == 10,
>>                "dvsec header size incorrect");
>> 
>> +/* CXL 2.0 - 8.1.11 */
>> +
>> /*
>>  * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
>>  * implement others.

Re: [RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Jonathan Cameron]

On Tue, 9 Feb 2021 15:36:03 -0500
Chris Browy <cbrowy@avery-design.com> wrote:

Split this into two patches for v3.  CDAT in one, compliance mode in the other.

I'd also move the actual elements out into the cxl components so that we
can register only what makes sense for a given device.   My guess
is that for now that will be static const anyway.

Coming together fine. Hopefully I'll start poking at the linux side of things
next week.  First job being simply providing a file to allow us to dump
the whole CDAT table.  Let me know if you get this loading an .aml file
in the meantime as that'll make it easier to test (if not I'll hack it
on top of these patches)

If needed I'll add it to iASL as well (may well be already in hand!)

I think my version of this stuff did a useful job in improving my understanding
of what we were trying to do, but that done I'm assuming we'll just abandon it
as the disposable prototype it was :)

Jonathan


> ---
>  hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
>  hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
>  include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
>  include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
>  include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
>  include/hw/cxl/cxl_device.h    |   3 +
>  include/hw/cxl/cxl_pci.h       |   4 +
>  7 files changed, 846 insertions(+)
>  create mode 100644 include/hw/cxl/cxl_cdat.h
>  create mode 100644 include/hw/cxl/cxl_compl.h
> 
> diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
> index e1bcee5..fc6c538 100644
> --- a/hw/cxl/cxl-component-utils.c
> +++ b/hw/cxl/cxl-component-utils.c
> @@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
>      range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
>      cxl->dvsec_offset += length;
>  }
> +
> +/* Return the sum of bytes */
> +static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
> +{
> +    cs->base = base;
> +    cs->length = len;
> +}
> +
> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
> +{
> +    uint8_t sum = 0;
> +    uint32_t len = 0;
> +    int i, j;
> +
> +    cxl_cstate->cdat_ent_len = 7;
> +    cxl_cstate->cdat_ent =
> +        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
> +
> +    cdat_ent_init(&cxl_cstate->cdat_ent[0],
> +                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[1],
> +                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[2],
> +                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[3],
> +                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[4],
> +                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[5],
> +                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
> +    cdat_ent_init(&cxl_cstate->cdat_ent[6],
> +                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
> +
> +    /* Set the DSMAS entry, ent = 1 */
> +    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
> +    cxl_cstate->dsmas.header.reserved = 0x0;
> +    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
> +    cxl_cstate->dsmas.DSMADhandle = 0x0;
> +    cxl_cstate->dsmas.flags = 0x0;
> +    cxl_cstate->dsmas.reserved2 = 0x0;
> +    cxl_cstate->dsmas.DPA_base = 0x0;
> +    cxl_cstate->dsmas.DPA_length = 0x40000;

Look to move the instances of these down into the memory device and expose
cdat_ent_init() to there.

That way, we can add whatever elements make sense for each type
of component.

Also have a cdat_ents_finalize() or similar to call at the end
which calculates overall length + checksum.

Should also be easy enough to add a simple bit of code to call
cdat_ent_init() for each element of a passed in CDAT.aml file.

> +
> +    /* Set the DSLBIS entry, ent = 2 */
> +    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
> +    cxl_cstate->dslbis.header.reserved = 0;
> +    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
> +    cxl_cstate->dslbis.handle = 0;
> +    cxl_cstate->dslbis.flags = 0;
> +    cxl_cstate->dslbis.data_type = 0;
> +    cxl_cstate->dslbis.reserved2 = 0;
> +    cxl_cstate->dslbis.entry_base_unit = 0;
> +    cxl_cstate->dslbis.entry[0] = 0;
> +    cxl_cstate->dslbis.entry[1] = 0;
> +    cxl_cstate->dslbis.entry[2] = 0;
> +    cxl_cstate->dslbis.reserved3 = 0;
> +
> +    /* Set the DSMSCIS entry, ent = 3 */
> +    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
> +    cxl_cstate->dsmscis.header.reserved = 0;
> +    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
> +    cxl_cstate->dsmscis.DSMASH_handle = 0;
> +    cxl_cstate->dsmscis.reserved2[0] = 0;
> +    cxl_cstate->dsmscis.reserved2[1] = 0;
> +    cxl_cstate->dsmscis.reserved2[2] = 0;
> +    cxl_cstate->dsmscis.memory_side_cache_size = 0;
> +    cxl_cstate->dsmscis.cache_attributes = 0;
> +
> +    /* Set the DSIS entry, ent = 4 */
> +    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
> +    cxl_cstate->dsis.header.reserved = 0;
> +    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
> +    cxl_cstate->dsis.flags = 0;
> +    cxl_cstate->dsis.handle = 0;
> +    cxl_cstate->dsis.reserved2 = 0;
> +
> +    /* Set the DSEMTS entry, ent = 5 */
> +    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
> +    cxl_cstate->dsemts.header.reserved = 0;
> +    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
> +    cxl_cstate->dsemts.DSMAS_handle = 0;
> +    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
> +    cxl_cstate->dsemts.reserved2 = 0;
> +    cxl_cstate->dsemts.DPA_offset = 0;
> +    cxl_cstate->dsemts.DPA_length = 0;
> +
> +    /* Set the SSLBIS entry, ent = 6 */
> +    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
> +    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
> +    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
> +    cxl_cstate->sslbis.sslbis_h.data_type = 0;
> +    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
> +    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
> +    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
> +    /* Set the SSLBE entry */
> +    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
> +    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
> +    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
> +    cxl_cstate->sslbis.sslbe[0].reserved = 0;
> +    /* Set the SSLBE entry */
> +    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
> +    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
> +    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
> +    cxl_cstate->sslbis.sslbe[1].reserved = 0;
> +    /* Set the SSLBE entry */
> +    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
> +    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
> +    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
> +    cxl_cstate->sslbis.sslbe[2].reserved = 0;
> +
> +    /* Set CDAT header, ent = 0 */
> +    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
> +    cxl_cstate->cdat_header.reserved[0] = 0;
> +    cxl_cstate->cdat_header.reserved[1] = 0;
> +    cxl_cstate->cdat_header.reserved[2] = 0;
> +    cxl_cstate->cdat_header.reserved[3] = 0;
> +    cxl_cstate->cdat_header.reserved[4] = 0;
> +    cxl_cstate->cdat_header.reserved[5] = 0;
> +    cxl_cstate->cdat_header.sequence = 0;
> +
> +    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
> +        /* Add length of each CDAT struct to total length */
> +        len = cxl_cstate->cdat_ent[i].length;
> +        cxl_cstate->cdat_header.length += len;
> +
> +        /* Calculate checksum of each CDAT struct */
> +        for (j = 0; j < len; j++) {
> +            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
> +        }
> +    }
> +    cxl_cstate->cdat_header.checksum = ~sum + 1;
> +}
> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
> index d091e64..86c762f 100644
> --- a/hw/mem/cxl_type3.c
> +++ b/hw/mem/cxl_type3.c
> @@ -13,6 +13,150 @@
>  #include "qemu/rcu.h"
>  #include "sysemu/hostmem.h"
>  #include "hw/cxl/cxl.h"
> +#include "hw/pci/msi.h"
> +#include "hw/pci/msix.h"
> +
> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap)
local to file, static and remove from header.
> +{
> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> +    uint32_t req;
> +    uint32_t byte_cnt = 0;
> +
> +    DOE_DBG(">> %s\n",  __func__);
> +
> +    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
> +        ->req_code;
> +    switch (req) {
> +    case CXL_COMP_MODE_CAP:
> +        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);

Use a local variable to cap_rsp or assign it via a structure here.
Basically get rid of the repitition of
cxl_cstate->doe_resp.cap_rsp
in the interests of readability.


> +        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
> +        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
> +        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
> +        cxl_cstate->doe_resp.cap_rsp.header.length =
> +            dwsizeof(struct cxl_compliance_mode_cap_rsp);
> +        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
> +        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
> +        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
> +        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
> +        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
> +        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
> +        break;
> +    case CXL_COMP_MODE_STATUS:
> +        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
> +        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
> +        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
> +        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
> +        cxl_cstate->doe_resp.status_rsp.header.length =
> +            dwsizeof(struct cxl_compliance_mode_status_rsp);
> +        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
> +        cxl_cstate->doe_resp.status_rsp.version = 0x1;
> +        cxl_cstate->doe_resp.status_rsp.length = 0x14;
> +        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
> +        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
> +        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
> +        break;
> +    default:

I guess the intent at somepoint is to actually hook some of these up?

> +        break;
> +    }
> +
> +    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
> +    DOE_DBG("<< %s\n",  __func__);
> +    return byte_cnt;
> +}
> +
> +
> +bool cxl_doe_compliance_rsp(DOECap *doe_cap)

Currently this is local to this file, so drop it form the header and
mark it static.  

> +{
> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> +    uint32_t byte_cnt;
> +    uint32_t dw_cnt;
> +
> +    DOE_DBG(">> %s\n",  __func__);
> +
> +    byte_cnt = cxl_doe_compliance_init(doe_cap);
> +    dw_cnt = byte_cnt / 4;
> +    memcpy(doe_cap->read_mbox,
> +           cxl_cstate->doe_resp.data_byte, byte_cnt);
> +
> +    doe_cap->read_mbox_len += dw_cnt;
> +
> +    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
> +            doe_cap->read_mbox_len,
> +            *(doe_cap->read_mbox + dw_cnt - 1));
> +
> +    DOE_DBG("<< %s\n",  __func__);
> +
> +    return DOE_SUCCESS;
> +}
> +
> +bool cxl_doe_cdat_rsp(DOECap *doe_cap)
Local to this file I think so drop it from header and mark it static.

> +{
> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> +    uint16_t ent;
> +    void *base;
> +    uint32_t len;
> +    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
> +
> +    ent = req->entry_handle;
> +    base = cxl_cstate->cdat_ent[ent].base;
> +    len = cxl_cstate->cdat_ent[ent].length;
> +
> +    struct cxl_cdat_rsp rsp = {
> +        .header = {
> +            .vendor_id = CXL_VENDOR_ID,
> +            .doe_type = CXL_DOE_TABLE_ACCESS,
> +            .reserved = 0x0,
> +            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
> +        },
> +        .req_code = CXL_DOE_TAB_RSP,
> +        .table_type = CXL_DOE_TAB_TYPE_CDAT,
> +        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
> +    };
> +
> +    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
> +    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
> +
> +    doe_cap->read_mbox_len += rsp.header.length;
> +    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
> +
> +    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
> +        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
> +    }
> +
> +    return DOE_SUCCESS;
> +}
> +
> +static uint32_t ct3d_config_read(PCIDevice *pci_dev,
> +                            uint32_t addr, int size)
> +{
> +    CXLType3Dev *ct3d = CT3(pci_dev);
> +    PCIEDOE *doe = &ct3d->doe;
> +    DOECap *doe_cap;
> +
> +    doe_cap = pcie_doe_covers_addr(doe, addr);
> +    if (doe_cap) {
> +        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
> +        return pcie_doe_read_config(doe_cap, addr, size);
> +    } else {
> +        return pci_default_read_config(pci_dev, addr, size);
> +    }
> +}
> +
> +static void ct3d_config_write(PCIDevice *pci_dev,
> +                            uint32_t addr, uint32_t val, int size)
> +{
> +    CXLType3Dev *ct3d = CT3(pci_dev);
> +    PCIEDOE *doe = &ct3d->doe;
> +    DOECap *doe_cap;
> +
> +    doe_cap = pcie_doe_covers_addr(doe, addr);
> +    if (doe_cap) {
> +        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
> +        pcie_doe_write_config(doe_cap, addr, val, size);
> +    } else {
> +        pci_default_write_config(pci_dev, addr, val, size);
> +    }
> +}
>  
>  static void build_dvsecs(CXLType3Dev *ct3d)
>  {
> @@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>      ComponentRegisters *regs = &cxl_cstate->crb;
>      MemoryRegion *mr = &regs->component_registers;
>      uint8_t *pci_conf = pci_dev->config;
> +    unsigned short msix_num = 2;
> +    int rc;
> +    int i;
>  
>      if (!ct3d->cxl_dstate.pmem) {
>          cxl_setup_memory(ct3d, errp);
> @@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>                       PCI_BASE_ADDRESS_SPACE_MEMORY |
>                           PCI_BASE_ADDRESS_MEM_TYPE_64,
>                       &ct3d->cxl_dstate.device_registers);
> +
> +    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
> +    for (i = 0; i < msix_num; i++) {
> +        msix_vector_use(pci_dev, i);
> +    }
> +
> +    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */

Tidy this up or parameterize it.

> +
> +    pcie_doe_init(pci_dev, &ct3d->doe);
> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);

check rc here.

> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
> +    if (rc) {
> +        error_setg(errp, "fail to add DOE cap");
> +        return;
> +    }
> +
> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
> +                               cxl_doe_compliance_rsp);
> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
> +                               cxl_doe_cdat_rsp);
> +
> +    cxl_doe_cdat_init(cxl_cstate);
>  }
>  
>  static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
> @@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
>      DeviceClass *dc = DEVICE_CLASS(oc);
>      PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
>      MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
> +
> +    pc->config_write = ct3d_config_write;
> +    pc->config_read = ct3d_config_read;
>      CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
>  
>      pc->realize = ct3_realize;
> diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
> new file mode 100644
> index 0000000..fbbd494
> --- /dev/null
> +++ b/include/hw/cxl/cxl_cdat.h
> @@ -0,0 +1,120 @@
> +#include "hw/cxl/cxl_pci.h"
> +
> +
> +enum {
> +    CXL_DOE_COMPLIANCE             = 0,
> +    CXL_DOE_TABLE_ACCESS           = 2,
> +    CXL_DOE_MAX_PROTOCOL
> +};
> +
> +#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
> +#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
> +
> +/*
> + * DOE CDAT Table Protocol (CXL Spec)
> + */
> +#define CXL_DOE_TAB_REQ 0
> +#define CXL_DOE_TAB_RSP 0
> +#define CXL_DOE_TAB_TYPE_CDAT 0
> +#define CXL_DOE_TAB_ENT_MAX 0xFFFF
> +
> +/* Read Entry Request, 8.1.11.1 Table 134 */
> +struct cxl_cdat {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t table_type;
> +    uint16_t entry_handle;
> +} QEMU_PACKED;
> +
> +/* Read Entry Response, 8.1.11.1 Table 135 */
> +#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
> +
... Note I'm just snipping out these big defines as I check them
against the spec :) Hence the gap.
...

> +struct cdat_dsmscis {
> +    struct cdat_sub_header header;
> +    uint8_t DSMASH_handle;

DMSAS_handle;

> +    uint8_t reserved2[3];
> +    uint64_t memory_side_cache_size;
> +    uint32_t cache_attributes;
> +} QEMU_PACKED;
> +


> +
> +struct cdat_sslbis_header {
> +    struct cdat_sub_header header;
> +    uint8_t data_type;
> +    uint8_t reserved2[3];
> +    uint64_t entry_base_unit;
> +} QEMU_PACKED;
> diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
> new file mode 100644
> index 0000000..ebbe488
> --- /dev/null
> +++ b/include/hw/cxl/cxl_compl.h
> @@ -0,0 +1,289 @@
> +/*
> + * CXL Compliance Mode Protocol

Needs license etc I think

> + */

A bunch of the responses below are all of this form, perhaps we
could just have one cxl_compliance_mode_generic_status_rsp ?
if you really want to then define the others perhaps use
#define to do it rather than lots of identical structures
each specified fully.


> +struct cxl_compliance_mode_inject_mac_delay_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t length;
> +    uint8_t status;
> +} QEMU_PACKED;
> +



> +struct cxl_compliance_mode_ignore_bit_error {
> +    DOEHeader header;
> +    uint8_t req_code;
> +    uint8_t version;
> +    uint16_t reserved;
> +    uint8_t opcode;
> +} QEMU_PACKED;
> +
This last one doesn't seem to line up with the CXL 2.0 spec.

> +struct cxl_compliance_mode_ignore_bit_error_rsp {
> +    DOEHeader header;
> +    uint8_t rsp_code;
> +    uint8_t version;
> +    uint8_t reserved[6];
> +} QEMU_PACKED;


> diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
> index 762feb5..23923df 100644
> --- a/include/hw/cxl/cxl_component.h
> +++ b/include/hw/cxl/cxl_component.h
> @@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
>  _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
>                 "No space for registers");

...

>  typedef struct component_registers {
>      /*
>       * Main memory region to be registered with QEMU core.
> @@ -160,6 +177,10 @@ typedef struct component_registers {
>      MemoryRegionOps *special_ops;
>  } ComponentRegisters;
>  
> +typedef struct cdat_struct {
> +    void *base;
> +    uint32_t length;
> +} CDATStruct;
>  /*
>   * A CXL component represents all entities in a CXL hierarchy. This includes,
>   * host bridges, root ports, upstream/downstream switch ports, and devices
> @@ -173,6 +194,104 @@ typedef struct cxl_component {
>              struct PCIDevice *pdev;
>          };
>      };
> +
> +    /*
> +     * SW write write mailbox and GO on last DW
> +     * device sets READY of offset DW in DO types to copy
> +     * to read mailbox register on subsequent cfg_read
> +     * of read mailbox register and then on cfg_write to
> +     * denote success read increments offset to next DW
> +     */
> +
> +    union doe_request_u {
> +        /* Compliance DOE Data Objects Type=0*/
> +        struct cxl_compliance_mode_cap
> +            mode_cap;

I'd only add line breaks for the longer ones of these.

> +        struct cxl_compliance_mode_status
> +            mode_status;
> +        struct cxl_compliance_mode_halt
> +            mode_halt;
> +        struct cxl_compliance_mode_multiple_write_streaming
> +            multiple_write_streaming;
> +        struct cxl_compliance_mode_producer_consumer
> +            producer_consumer;
> +        struct cxl_compliance_mode_inject_bogus_writes
> +            inject_bogus_writes;
> +        struct cxl_compliance_mode_inject_poison
> +            inject_poison;
> +        struct cxl_compliance_mode_inject_crc
> +            inject_crc;
> +        struct cxl_compliance_mode_inject_flow_control
> +            inject_flow_control;
> +        struct cxl_compliance_mode_toggle_cache_flush
> +            toggle_cache_flush;
> +        struct cxl_compliance_mode_inject_mac_delay
> +            inject_mac_delay;
> +        struct cxl_compliance_mode_insert_unexp_mac
> +            insert_unexp_mac;
> +        struct cxl_compliance_mode_inject_viral
> +            inject_viral;
> +        struct cxl_compliance_mode_inject_almp
> +            inject_almp;
> +        struct cxl_compliance_mode_ignore_almp
> +            ignore_almp;
> +        struct cxl_compliance_mode_ignore_bit_error
> +            ignore_bit_error;
> +        char data_byte[128];
> +    } doe_request;
> +
> +    union doe_resp_u {
> +        /* Compliance DOE Data Objects Type=0*/
> +        struct cxl_compliance_mode_cap_rsp
> +            cap_rsp;
> +        struct cxl_compliance_mode_status_rsp
> +            status_rsp;
> +        struct cxl_compliance_mode_halt_rsp
> +            halt_rsp;
> +        struct cxl_compliance_mode_multiple_write_streaming_rsp
> +            multiple_write_streaming_rsp;
> +        struct cxl_compliance_mode_producer_consumer_rsp
> +            producer_consumer_rsp;
> +        struct cxl_compliance_mode_inject_bogus_writes_rsp
> +            inject_bogus_writes_rsp;
> +        struct cxl_compliance_mode_inject_poison_rsp
> +            inject_poison_rsp;
> +        struct cxl_compliance_mode_inject_crc_rsp
> +            inject_crc_rsp;
> +        struct cxl_compliance_mode_inject_flow_control_rsp
> +            inject_flow_control_rsp;
> +        struct cxl_compliance_mode_toggle_cache_flush_rsp
> +            toggle_cache_flush_rsp;
> +        struct cxl_compliance_mode_inject_mac_delay_rsp
> +            inject_mac_delay_rsp;
> +        struct cxl_compliance_mode_insert_unexp_mac_rsp
> +            insert_unexp_mac_rsp;
> +        struct cxl_compliance_mode_inject_viral
> +            inject_viral_rsp;
> +        struct cxl_compliance_mode_inject_almp_rsp
> +            inject_almp_rsp;
> +        struct cxl_compliance_mode_ignore_almp_rsp
> +            ignore_almp_rsp;
> +        struct cxl_compliance_mode_ignore_bit_error_rsp
> +            ignore_bit_error_rsp;
> +        char data_byte[520 * 4];
> +        uint32_t data_dword[520];
> +    } doe_resp;
> +
> +    /* Table entry types */
Hmm. Not sure all CXL components will have CDAT.  Root ports for
example?

> +    struct cdat_table_header cdat_header;
> +    struct cdat_dsmas dsmas;
> +    struct cdat_dslbis dslbis;

As I said in v1, some of these will need to be multiples so this
flat structure just doesn't work.

> +    struct cdat_dsmscis dsmscis;
> +    struct cdat_dsis dsis;
> +    struct cdat_dsemts dsemts;
> +    struct {
> +        struct cdat_sslbis_header sslbis_h;
> +        struct cdat_sslbe sslbe[3];

I'm curious.  Why 3?  This is between pairs of ports so 1USP 2DSP switch?

> +    } sslbis;


> +
> +    CDATStruct *cdat_ent;
> +    int cdat_ent_len;
>  } CXLComponentState;
>  
>  void cxl_component_register_block_init(Object *obj,
> @@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
>  void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
>                                  uint16_t type, uint8_t rev, uint8_t *body);
>  
> +void cxl_component_create_doe(CXLComponentState *cxl_cstate,
> +                              uint16_t offset, unsigned vec);

Doesn't seem to exist.

Some of the following are local to one file so shouldn't be here iether.

> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
> +bool cxl_doe_compliance_rsp(DOECap *doe_cap);
> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
> +bool cxl_doe_cdat_rsp(DOECap *doe_cap);
> +uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);

Doesn't seem to exist.

>  #endif
> diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
> index c608ced..08bf646 100644
> --- a/include/hw/cxl/cxl_device.h
> +++ b/include/hw/cxl/cxl_device.h
> @@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
>      /* State */
>      CXLComponentState cxl_cstate;
>      CXLDeviceState cxl_dstate;
> +
> +    /* DOE */
> +    PCIEDOE doe;
>  } CXLType3Dev;
>  
>  #ifndef TYPE_CXL_TYPE3_DEV
> diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
> index 9ec28c9..5cab197 100644
> --- a/include/hw/cxl/cxl_pci.h
> +++ b/include/hw/cxl/cxl_pci.h
> @@ -12,6 +12,8 @@
>  
>  #include "hw/pci/pci.h"
>  #include "hw/pci/pcie.h"
> +#include "hw/cxl/cxl_cdat.h"
> +#include "hw/cxl/cxl_compl.h"
>  
>  #define CXL_VENDOR_ID 0x1e98
>  
> @@ -54,6 +56,8 @@ struct dvsec_header {
>  _Static_assert(sizeof(struct dvsec_header) == 10,
>                 "dvsec header size incorrect");
>  
> +/* CXL 2.0 - 8.1.11 */
> +

Clean this out next time - doesn't belong in this patch.

>  /*
>   * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
>   * implement others.

Re: [RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Chris Browy]

> On Feb 12, 2021, at 12:23 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
> 
> On Tue, 9 Feb 2021 15:36:03 -0500
> Chris Browy <cbrowy@avery-design.com> wrote:
> 
> Split this into two patches for v3.  CDAT in one, compliance mode in the other.
> 

Compliance mode is an optional feature.  We’ll split it out.

> I'd also move the actual elements out into the cxl components so that we
> can register only what makes sense for a given device.   My guess
> is that for now that will be static const anyway.
> 
> Coming together fine. Hopefully I'll start poking at the linux side of things
> next week.  First job being simply providing a file to allow us to dump
> the whole CDAT table.  Let me know if you get this loading an .aml file
> in the meantime as that'll make it easier to test (if not I'll hack it
> on top of these patches)

We can get the .aml loading by Thurs next week.  Holiday next few days for 
some of our folks.

> 
> If needed I'll add it to iASL as well (may well be already in hand!)
> 
> I think my version of this stuff did a useful job in improving my understanding
> of what we were trying to do, but that done I'm assuming we'll just abandon it
> as the disposable prototype it was :)
> 

Thanks for focusing in on the area and uncovering problems with both our versions!

Still lots of pieces need to come together and get working to be able to fully enumerate 
and configure the device!

> Jonathan
> 
> 
>> ---
>> hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
>> hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
>> include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
>> include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
>> include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
>> include/hw/cxl/cxl_device.h    |   3 +
>> include/hw/cxl/cxl_pci.h       |   4 +
>> 7 files changed, 846 insertions(+)
>> create mode 100644 include/hw/cxl/cxl_cdat.h
>> create mode 100644 include/hw/cxl/cxl_compl.h
>> 
>> diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
>> index e1bcee5..fc6c538 100644
>> --- a/hw/cxl/cxl-component-utils.c
>> +++ b/hw/cxl/cxl-component-utils.c
>> @@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
>>     range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
>>     cxl->dvsec_offset += length;
>> }
>> +
>> +/* Return the sum of bytes */
>> +static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
>> +{
>> +    cs->base = base;
>> +    cs->length = len;
>> +}
>> +
>> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
>> +{
>> +    uint8_t sum = 0;
>> +    uint32_t len = 0;
>> +    int i, j;
>> +
>> +    cxl_cstate->cdat_ent_len = 7;
>> +    cxl_cstate->cdat_ent =
>> +        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
>> +
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[0],
>> +                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[1],
>> +                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[2],
>> +                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[3],
>> +                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[4],
>> +                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[5],
>> +                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
>> +    cdat_ent_init(&cxl_cstate->cdat_ent[6],
>> +                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
>> +
>> +    /* Set the DSMAS entry, ent = 1 */
>> +    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
>> +    cxl_cstate->dsmas.header.reserved = 0x0;
>> +    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
>> +    cxl_cstate->dsmas.DSMADhandle = 0x0;
>> +    cxl_cstate->dsmas.flags = 0x0;
>> +    cxl_cstate->dsmas.reserved2 = 0x0;
>> +    cxl_cstate->dsmas.DPA_base = 0x0;
>> +    cxl_cstate->dsmas.DPA_length = 0x40000;
> 
> Look to move the instances of these down into the memory device and expose
> cdat_ent_init() to there.
> 
> That way, we can add whatever elements make sense for each type
> of component.
>  

> Also have a cdat_ents_finalize() or similar to call at the end
> which calculates overall length + checksum.

OK

> 
> Should also be easy enough to add a simple bit of code to call
> cdat_ent_init() for each element of a passed in CDAT.aml file.
> 

We’ll address all the above when we add the CDAT.aml file support 
which may only pass a subset of structures. 

>> +
>> +    /* Set the DSLBIS entry, ent = 2 */
>> +    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
>> +    cxl_cstate->dslbis.header.reserved = 0;
>> +    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
>> +    cxl_cstate->dslbis.handle = 0;
>> +    cxl_cstate->dslbis.flags = 0;
>> +    cxl_cstate->dslbis.data_type = 0;
>> +    cxl_cstate->dslbis.reserved2 = 0;
>> +    cxl_cstate->dslbis.entry_base_unit = 0;
>> +    cxl_cstate->dslbis.entry[0] = 0;
>> +    cxl_cstate->dslbis.entry[1] = 0;
>> +    cxl_cstate->dslbis.entry[2] = 0;
>> +    cxl_cstate->dslbis.reserved3 = 0;
>> +
>> +    /* Set the DSMSCIS entry, ent = 3 */
>> +    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
>> +    cxl_cstate->dsmscis.header.reserved = 0;
>> +    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
>> +    cxl_cstate->dsmscis.DSMASH_handle = 0;
>> +    cxl_cstate->dsmscis.reserved2[0] = 0;
>> +    cxl_cstate->dsmscis.reserved2[1] = 0;
>> +    cxl_cstate->dsmscis.reserved2[2] = 0;
>> +    cxl_cstate->dsmscis.memory_side_cache_size = 0;
>> +    cxl_cstate->dsmscis.cache_attributes = 0;
>> +
>> +    /* Set the DSIS entry, ent = 4 */
>> +    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
>> +    cxl_cstate->dsis.header.reserved = 0;
>> +    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
>> +    cxl_cstate->dsis.flags = 0;
>> +    cxl_cstate->dsis.handle = 0;
>> +    cxl_cstate->dsis.reserved2 = 0;
>> +
>> +    /* Set the DSEMTS entry, ent = 5 */
>> +    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
>> +    cxl_cstate->dsemts.header.reserved = 0;
>> +    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
>> +    cxl_cstate->dsemts.DSMAS_handle = 0;
>> +    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
>> +    cxl_cstate->dsemts.reserved2 = 0;
>> +    cxl_cstate->dsemts.DPA_offset = 0;
>> +    cxl_cstate->dsemts.DPA_length = 0;
>> +
>> +    /* Set the SSLBIS entry, ent = 6 */
>> +    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
>> +    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
>> +    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
>> +    cxl_cstate->sslbis.sslbis_h.data_type = 0;
>> +    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
>> +    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
>> +    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
>> +    /* Set the SSLBE entry */
>> +    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
>> +    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
>> +    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
>> +    cxl_cstate->sslbis.sslbe[0].reserved = 0;
>> +    /* Set the SSLBE entry */
>> +    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
>> +    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
>> +    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
>> +    cxl_cstate->sslbis.sslbe[1].reserved = 0;
>> +    /* Set the SSLBE entry */
>> +    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
>> +    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
>> +    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
>> +    cxl_cstate->sslbis.sslbe[2].reserved = 0;
>> +
>> +    /* Set CDAT header, ent = 0 */
>> +    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
>> +    cxl_cstate->cdat_header.reserved[0] = 0;
>> +    cxl_cstate->cdat_header.reserved[1] = 0;
>> +    cxl_cstate->cdat_header.reserved[2] = 0;
>> +    cxl_cstate->cdat_header.reserved[3] = 0;
>> +    cxl_cstate->cdat_header.reserved[4] = 0;
>> +    cxl_cstate->cdat_header.reserved[5] = 0;
>> +    cxl_cstate->cdat_header.sequence = 0;
>> +
>> +    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
>> +        /* Add length of each CDAT struct to total length */
>> +        len = cxl_cstate->cdat_ent[i].length;
>> +        cxl_cstate->cdat_header.length += len;
>> +
>> +        /* Calculate checksum of each CDAT struct */
>> +        for (j = 0; j < len; j++) {
>> +            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
>> +        }
>> +    }
>> +    cxl_cstate->cdat_header.checksum = ~sum + 1;
>> +}
>> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
>> index d091e64..86c762f 100644
>> --- a/hw/mem/cxl_type3.c
>> +++ b/hw/mem/cxl_type3.c
>> @@ -13,6 +13,150 @@
>> #include "qemu/rcu.h"
>> #include "sysemu/hostmem.h"
>> #include "hw/cxl/cxl.h"
>> +#include "hw/pci/msi.h"
>> +#include "hw/pci/msix.h"
>> +
>> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap)
> local to file, static and remove from header.
>> +{
>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>> +    uint32_t req;
>> +    uint32_t byte_cnt = 0;
>> +
>> +    DOE_DBG(">> %s\n",  __func__);
>> +
>> +    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
>> +        ->req_code;
>> +    switch (req) {
>> +    case CXL_COMP_MODE_CAP:
>> +        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);
> 
> Use a local variable to cap_rsp or assign it via a structure here.
> Basically get rid of the repitition of
> cxl_cstate->doe_resp.cap_rsp
> in the interests of readability.
> 
> 
>> +        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
>> +        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
>> +        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
>> +        cxl_cstate->doe_resp.cap_rsp.header.length =
>> +            dwsizeof(struct cxl_compliance_mode_cap_rsp);
>> +        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
>> +        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
>> +        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
>> +        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
>> +        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
>> +        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
>> +        break;
>> +    case CXL_COMP_MODE_STATUS:
>> +        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
>> +        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
>> +        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
>> +        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
>> +        cxl_cstate->doe_resp.status_rsp.header.length =
>> +            dwsizeof(struct cxl_compliance_mode_status_rsp);
>> +        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
>> +        cxl_cstate->doe_resp.status_rsp.version = 0x1;
>> +        cxl_cstate->doe_resp.status_rsp.length = 0x14;
>> +        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
>> +        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
>> +        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
>> +        break;
>> +    default:
> 
> I guess the intent at somepoint is to actually hook some of these up?
> 
>> +        break;
>> +    }
>> +
>> +    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
>> +    DOE_DBG("<< %s\n",  __func__);
>> +    return byte_cnt;
>> +}
>> +
>> +
>> +bool cxl_doe_compliance_rsp(DOECap *doe_cap)
> 
> Currently this is local to this file, so drop it form the header and
> mark it static.  
> 
>> +{
>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>> +    uint32_t byte_cnt;
>> +    uint32_t dw_cnt;
>> +
>> +    DOE_DBG(">> %s\n",  __func__);
>> +
>> +    byte_cnt = cxl_doe_compliance_init(doe_cap);
>> +    dw_cnt = byte_cnt / 4;
>> +    memcpy(doe_cap->read_mbox,
>> +           cxl_cstate->doe_resp.data_byte, byte_cnt);
>> +
>> +    doe_cap->read_mbox_len += dw_cnt;
>> +
>> +    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
>> +            doe_cap->read_mbox_len,
>> +            *(doe_cap->read_mbox + dw_cnt - 1));
>> +
>> +    DOE_DBG("<< %s\n",  __func__);
>> +
>> +    return DOE_SUCCESS;
>> +}
>> +
>> +bool cxl_doe_cdat_rsp(DOECap *doe_cap)
> Local to this file I think so drop it from header and mark it static.
> 
>> +{
>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>> +    uint16_t ent;
>> +    void *base;
>> +    uint32_t len;
>> +    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
>> +
>> +    ent = req->entry_handle;
>> +    base = cxl_cstate->cdat_ent[ent].base;
>> +    len = cxl_cstate->cdat_ent[ent].length;
>> +
>> +    struct cxl_cdat_rsp rsp = {
>> +        .header = {
>> +            .vendor_id = CXL_VENDOR_ID,
>> +            .doe_type = CXL_DOE_TABLE_ACCESS,
>> +            .reserved = 0x0,
>> +            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
>> +        },
>> +        .req_code = CXL_DOE_TAB_RSP,
>> +        .table_type = CXL_DOE_TAB_TYPE_CDAT,
>> +        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
>> +    };
>> +
>> +    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
>> +    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
>> +
>> +    doe_cap->read_mbox_len += rsp.header.length;
>> +    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
>> +
>> +    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
>> +        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
>> +    }
>> +
>> +    return DOE_SUCCESS;
>> +}
>> +
>> +static uint32_t ct3d_config_read(PCIDevice *pci_dev,
>> +                            uint32_t addr, int size)
>> +{
>> +    CXLType3Dev *ct3d = CT3(pci_dev);
>> +    PCIEDOE *doe = &ct3d->doe;
>> +    DOECap *doe_cap;
>> +
>> +    doe_cap = pcie_doe_covers_addr(doe, addr);
>> +    if (doe_cap) {
>> +        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
>> +        return pcie_doe_read_config(doe_cap, addr, size);
>> +    } else {
>> +        return pci_default_read_config(pci_dev, addr, size);
>> +    }
>> +}
>> +
>> +static void ct3d_config_write(PCIDevice *pci_dev,
>> +                            uint32_t addr, uint32_t val, int size)
>> +{
>> +    CXLType3Dev *ct3d = CT3(pci_dev);
>> +    PCIEDOE *doe = &ct3d->doe;
>> +    DOECap *doe_cap;
>> +
>> +    doe_cap = pcie_doe_covers_addr(doe, addr);
>> +    if (doe_cap) {
>> +        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
>> +        pcie_doe_write_config(doe_cap, addr, val, size);
>> +    } else {
>> +        pci_default_write_config(pci_dev, addr, val, size);
>> +    }
>> +}
>> 
>> static void build_dvsecs(CXLType3Dev *ct3d)
>> {
>> @@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>>     ComponentRegisters *regs = &cxl_cstate->crb;
>>     MemoryRegion *mr = &regs->component_registers;
>>     uint8_t *pci_conf = pci_dev->config;
>> +    unsigned short msix_num = 2;
>> +    int rc;
>> +    int i;
>> 
>>     if (!ct3d->cxl_dstate.pmem) {
>>         cxl_setup_memory(ct3d, errp);
>> @@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>>                      PCI_BASE_ADDRESS_SPACE_MEMORY |
>>                          PCI_BASE_ADDRESS_MEM_TYPE_64,
>>                      &ct3d->cxl_dstate.device_registers);
>> +
>> +    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
>> +    for (i = 0; i < msix_num; i++) {
>> +        msix_vector_use(pci_dev, i);
>> +    }
>> +
>> +    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */
> 
> Tidy this up or parameterize it.
> 
>> +
>> +    pcie_doe_init(pci_dev, &ct3d->doe);
>> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);
> 
> check rc here.
> 
>> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
>> +    if (rc) {
>> +        error_setg(errp, "fail to add DOE cap");
>> +        return;
>> +    }
>> +
>> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
>> +                               cxl_doe_compliance_rsp);
>> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
>> +                               cxl_doe_cdat_rsp);
>> +
>> +    cxl_doe_cdat_init(cxl_cstate);
>> }
>> 
>> static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
>> @@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
>>     DeviceClass *dc = DEVICE_CLASS(oc);
>>     PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
>>     MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
>> +
>> +    pc->config_write = ct3d_config_write;
>> +    pc->config_read = ct3d_config_read;
>>     CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
>> 
>>     pc->realize = ct3_realize;
>> diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
>> new file mode 100644
>> index 0000000..fbbd494
>> --- /dev/null
>> +++ b/include/hw/cxl/cxl_cdat.h
>> @@ -0,0 +1,120 @@
>> +#include "hw/cxl/cxl_pci.h"
>> +
>> +
>> +enum {
>> +    CXL_DOE_COMPLIANCE             = 0,
>> +    CXL_DOE_TABLE_ACCESS           = 2,
>> +    CXL_DOE_MAX_PROTOCOL
>> +};
>> +
>> +#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
>> +#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
>> +
>> +/*
>> + * DOE CDAT Table Protocol (CXL Spec)
>> + */
>> +#define CXL_DOE_TAB_REQ 0
>> +#define CXL_DOE_TAB_RSP 0
>> +#define CXL_DOE_TAB_TYPE_CDAT 0
>> +#define CXL_DOE_TAB_ENT_MAX 0xFFFF
>> +
>> +/* Read Entry Request, 8.1.11.1 Table 134 */
>> +struct cxl_cdat {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t table_type;
>> +    uint16_t entry_handle;
>> +} QEMU_PACKED;
>> +
>> +/* Read Entry Response, 8.1.11.1 Table 135 */
>> +#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
>> +
> ... Note I'm just snipping out these big defines as I check them
> against the spec :) Hence the gap.
> ...
> 
>> +struct cdat_dsmscis {
>> +    struct cdat_sub_header header;
>> +    uint8_t DSMASH_handle;
> 
> DMSAS_handle;
> 
>> +    uint8_t reserved2[3];
>> +    uint64_t memory_side_cache_size;
>> +    uint32_t cache_attributes;
>> +} QEMU_PACKED;
>> +
> 
> 
>> +
>> +struct cdat_sslbis_header {
>> +    struct cdat_sub_header header;
>> +    uint8_t data_type;
>> +    uint8_t reserved2[3];
>> +    uint64_t entry_base_unit;
>> +} QEMU_PACKED;
>> diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
>> new file mode 100644
>> index 0000000..ebbe488
>> --- /dev/null
>> +++ b/include/hw/cxl/cxl_compl.h
>> @@ -0,0 +1,289 @@
>> +/*
>> + * CXL Compliance Mode Protocol
> 
> Needs license etc I think
> 
>> + */
> 
> A bunch of the responses below are all of this form, perhaps we
> could just have one cxl_compliance_mode_generic_status_rsp ?
> if you really want to then define the others perhaps use
> #define to do it rather than lots of identical structures
> each specified fully.


Got it.   Should have simplified it after cutting and pasting so many times.
> 
> 
>> +struct cxl_compliance_mode_inject_mac_delay_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t length;
>> +    uint8_t status;
>> +} QEMU_PACKED;
>> +
> 
> 
> 
>> +struct cxl_compliance_mode_ignore_bit_error {
>> +    DOEHeader header;
>> +    uint8_t req_code;
>> +    uint8_t version;
>> +    uint16_t reserved;
>> +    uint8_t opcode;
>> +} QEMU_PACKED;
>> +
> This last one doesn't seem to line up with the CXL 2.0 spec.

Good catch!

> 
>> +struct cxl_compliance_mode_ignore_bit_error_rsp {
>> +    DOEHeader header;
>> +    uint8_t rsp_code;
>> +    uint8_t version;
>> +    uint8_t reserved[6];
>> +} QEMU_PACKED;
> 
> 
>> diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
>> index 762feb5..23923df 100644
>> --- a/include/hw/cxl/cxl_component.h
>> +++ b/include/hw/cxl/cxl_component.h
>> @@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
>> _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
>>                "No space for registers");
> 
> ...
> 
>> typedef struct component_registers {
>>     /*
>>      * Main memory region to be registered with QEMU core.
>> @@ -160,6 +177,10 @@ typedef struct component_registers {
>>     MemoryRegionOps *special_ops;
>> } ComponentRegisters;
>> 
>> +typedef struct cdat_struct {
>> +    void *base;
>> +    uint32_t length;
>> +} CDATStruct;
>> /*
>>  * A CXL component represents all entities in a CXL hierarchy. This includes,
>>  * host bridges, root ports, upstream/downstream switch ports, and devices
>> @@ -173,6 +194,104 @@ typedef struct cxl_component {
>>             struct PCIDevice *pdev;
>>         };
>>     };
>> +
>> +    /*
>> +     * SW write write mailbox and GO on last DW
>> +     * device sets READY of offset DW in DO types to copy
>> +     * to read mailbox register on subsequent cfg_read
>> +     * of read mailbox register and then on cfg_write to
>> +     * denote success read increments offset to next DW
>> +     */
>> +
>> +    union doe_request_u {
>> +        /* Compliance DOE Data Objects Type=0*/
>> +        struct cxl_compliance_mode_cap
>> +            mode_cap;
> 
> I'd only add line breaks for the longer ones of these.
> 
>> +        struct cxl_compliance_mode_status
>> +            mode_status;
>> +        struct cxl_compliance_mode_halt
>> +            mode_halt;
>> +        struct cxl_compliance_mode_multiple_write_streaming
>> +            multiple_write_streaming;
>> +        struct cxl_compliance_mode_producer_consumer
>> +            producer_consumer;
>> +        struct cxl_compliance_mode_inject_bogus_writes
>> +            inject_bogus_writes;
>> +        struct cxl_compliance_mode_inject_poison
>> +            inject_poison;
>> +        struct cxl_compliance_mode_inject_crc
>> +            inject_crc;
>> +        struct cxl_compliance_mode_inject_flow_control
>> +            inject_flow_control;
>> +        struct cxl_compliance_mode_toggle_cache_flush
>> +            toggle_cache_flush;
>> +        struct cxl_compliance_mode_inject_mac_delay
>> +            inject_mac_delay;
>> +        struct cxl_compliance_mode_insert_unexp_mac
>> +            insert_unexp_mac;
>> +        struct cxl_compliance_mode_inject_viral
>> +            inject_viral;
>> +        struct cxl_compliance_mode_inject_almp
>> +            inject_almp;
>> +        struct cxl_compliance_mode_ignore_almp
>> +            ignore_almp;
>> +        struct cxl_compliance_mode_ignore_bit_error
>> +            ignore_bit_error;
>> +        char data_byte[128];
>> +    } doe_request;
>> +
>> +    union doe_resp_u {
>> +        /* Compliance DOE Data Objects Type=0*/
>> +        struct cxl_compliance_mode_cap_rsp
>> +            cap_rsp;
>> +        struct cxl_compliance_mode_status_rsp
>> +            status_rsp;
>> +        struct cxl_compliance_mode_halt_rsp
>> +            halt_rsp;
>> +        struct cxl_compliance_mode_multiple_write_streaming_rsp
>> +            multiple_write_streaming_rsp;
>> +        struct cxl_compliance_mode_producer_consumer_rsp
>> +            producer_consumer_rsp;
>> +        struct cxl_compliance_mode_inject_bogus_writes_rsp
>> +            inject_bogus_writes_rsp;
>> +        struct cxl_compliance_mode_inject_poison_rsp
>> +            inject_poison_rsp;
>> +        struct cxl_compliance_mode_inject_crc_rsp
>> +            inject_crc_rsp;
>> +        struct cxl_compliance_mode_inject_flow_control_rsp
>> +            inject_flow_control_rsp;
>> +        struct cxl_compliance_mode_toggle_cache_flush_rsp
>> +            toggle_cache_flush_rsp;
>> +        struct cxl_compliance_mode_inject_mac_delay_rsp
>> +            inject_mac_delay_rsp;
>> +        struct cxl_compliance_mode_insert_unexp_mac_rsp
>> +            insert_unexp_mac_rsp;
>> +        struct cxl_compliance_mode_inject_viral
>> +            inject_viral_rsp;
>> +        struct cxl_compliance_mode_inject_almp_rsp
>> +            inject_almp_rsp;
>> +        struct cxl_compliance_mode_ignore_almp_rsp
>> +            ignore_almp_rsp;
>> +        struct cxl_compliance_mode_ignore_bit_error_rsp
>> +            ignore_bit_error_rsp;
>> +        char data_byte[520 * 4];
>> +        uint32_t data_dword[520];
>> +    } doe_resp;
>> +
>> +    /* Table entry types */
> Hmm. Not sure all CXL components will have CDAT.  Root ports for
> example?
> 
>> +    struct cdat_table_header cdat_header;
>> +    struct cdat_dsmas dsmas;
>> +    struct cdat_dslbis dslbis;
> 
> As I said in v1, some of these will need to be multiples so this
> flat structure just doesn't work.
> 
>> +    struct cdat_dsmscis dsmscis;
>> +    struct cdat_dsis dsis;
>> +    struct cdat_dsemts dsemts;
>> +    struct {
>> +        struct cdat_sslbis_header sslbis_h;
>> +        struct cdat_sslbe sslbe[3];
> 
> I'm curious.  Why 3?  This is between pairs of ports so 1USP 2DSP switch?
> 
>> +    } sslbis;
> 
> 
>> +
>> +    CDATStruct *cdat_ent;
>> +    int cdat_ent_len;
>> } CXLComponentState;
>> 
>> void cxl_component_register_block_init(Object *obj,
>> @@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
>> void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
>>                                 uint16_t type, uint8_t rev, uint8_t *body);
>> 
>> +void cxl_component_create_doe(CXLComponentState *cxl_cstate,
>> +                              uint16_t offset, unsigned vec);
> 
> Doesn't seem to exist.
> 
> Some of the following are local to one file so shouldn't be here iether.
> 
>> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
>> +bool cxl_doe_compliance_rsp(DOECap *doe_cap);
>> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
>> +bool cxl_doe_cdat_rsp(DOECap *doe_cap);
>> +uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);
> 
> Doesn't seem to exist.
> 
>> #endif
>> diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
>> index c608ced..08bf646 100644
>> --- a/include/hw/cxl/cxl_device.h
>> +++ b/include/hw/cxl/cxl_device.h
>> @@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
>>     /* State */
>>     CXLComponentState cxl_cstate;
>>     CXLDeviceState cxl_dstate;
>> +
>> +    /* DOE */
>> +    PCIEDOE doe;
>> } CXLType3Dev;
>> 
>> #ifndef TYPE_CXL_TYPE3_DEV
>> diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
>> index 9ec28c9..5cab197 100644
>> --- a/include/hw/cxl/cxl_pci.h
>> +++ b/include/hw/cxl/cxl_pci.h
>> @@ -12,6 +12,8 @@
>> 
>> #include "hw/pci/pci.h"
>> #include "hw/pci/pcie.h"
>> +#include "hw/cxl/cxl_cdat.h"
>> +#include "hw/cxl/cxl_compl.h"
>> 
>> #define CXL_VENDOR_ID 0x1e98
>> 
>> @@ -54,6 +56,8 @@ struct dvsec_header {
>> _Static_assert(sizeof(struct dvsec_header) == 10,
>>                "dvsec header size incorrect");
>> 
>> +/* CXL 2.0 - 8.1.11 */
>> +
> 
> Clean this out next time - doesn't belong in this patch.
> 
>> /*
>>  * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
>>  * implement others.
> 

Re: [RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Jonathan Cameron]

On Fri, 12 Feb 2021 17:26:50 -0500
Chris Browy <cbrowy@avery-design.com> wrote:

> > On Feb 12, 2021, at 12:23 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
> > 
> > On Tue, 9 Feb 2021 15:36:03 -0500
> > Chris Browy <cbrowy@avery-design.com> wrote:
> > 
> > Split this into two patches for v3.  CDAT in one, compliance mode in the other.
> >   
> 
> Compliance mode is an optional feature.  We’ll split it out.
> 
> > I'd also move the actual elements out into the cxl components so that we
> > can register only what makes sense for a given device.   My guess
> > is that for now that will be static const anyway.
> > 
> > Coming together fine. Hopefully I'll start poking at the linux side of things
> > next week.  First job being simply providing a file to allow us to dump
> > the whole CDAT table.  Let me know if you get this loading an .aml file
> > in the meantime as that'll make it easier to test (if not I'll hack it
> > on top of these patches)  
> 
> We can get the .aml loading by Thurs next week.  Holiday next few days for 
> some of our folks.
> 
> > 
> > If needed I'll add it to iASL as well (may well be already in hand!)

There is a potential problem doing this.  CDAT doesn't have the table
type ID that an ACPI table would have.  That means raw CDAT tables
are not identifiable and I think this makes it hard to use iASL with them
without changing it's general means of functioning.

We can probably do something with an extra parameter, but this lack of
identifier is going to make it harder to persuade people that it's sensible to
including CDAT in iASL.

> > 
> > I think my version of this stuff did a useful job in improving my understanding
> > of what we were trying to do, but that done I'm assuming we'll just abandon it
> > as the disposable prototype it was :)
> >   
> 
> Thanks for focusing in on the area and uncovering problems with both our versions!
> 
> Still lots of pieces need to come together and get working to be able to fully enumerate 
> and configure the device!
> 
> > Jonathan
> > 
> >   
> >> ---
> >> hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
> >> hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
> >> include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
> >> include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
> >> include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
> >> include/hw/cxl/cxl_device.h    |   3 +
> >> include/hw/cxl/cxl_pci.h       |   4 +
> >> 7 files changed, 846 insertions(+)
> >> create mode 100644 include/hw/cxl/cxl_cdat.h
> >> create mode 100644 include/hw/cxl/cxl_compl.h
> >> 
> >> diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
> >> index e1bcee5..fc6c538 100644
> >> --- a/hw/cxl/cxl-component-utils.c
> >> +++ b/hw/cxl/cxl-component-utils.c
> >> @@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
> >>     range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
> >>     cxl->dvsec_offset += length;
> >> }
> >> +
> >> +/* Return the sum of bytes */
> >> +static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
> >> +{
> >> +    cs->base = base;
> >> +    cs->length = len;
> >> +}
> >> +
> >> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
> >> +{
> >> +    uint8_t sum = 0;
> >> +    uint32_t len = 0;
> >> +    int i, j;
> >> +
> >> +    cxl_cstate->cdat_ent_len = 7;
> >> +    cxl_cstate->cdat_ent =
> >> +        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
> >> +
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[0],
> >> +                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[1],
> >> +                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[2],
> >> +                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[3],
> >> +                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[4],
> >> +                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[5],
> >> +                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
> >> +    cdat_ent_init(&cxl_cstate->cdat_ent[6],
> >> +                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
> >> +
> >> +    /* Set the DSMAS entry, ent = 1 */
> >> +    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
> >> +    cxl_cstate->dsmas.header.reserved = 0x0;
> >> +    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
> >> +    cxl_cstate->dsmas.DSMADhandle = 0x0;
> >> +    cxl_cstate->dsmas.flags = 0x0;
> >> +    cxl_cstate->dsmas.reserved2 = 0x0;
> >> +    cxl_cstate->dsmas.DPA_base = 0x0;
> >> +    cxl_cstate->dsmas.DPA_length = 0x40000;  
> > 
> > Look to move the instances of these down into the memory device and expose
> > cdat_ent_init() to there.
> > 
> > That way, we can add whatever elements make sense for each type
> > of component.
> >    
> 
> > Also have a cdat_ents_finalize() or similar to call at the end
> > which calculates overall length + checksum.  
> 
> OK
> 
> > 
> > Should also be easy enough to add a simple bit of code to call
> > cdat_ent_init() for each element of a passed in CDAT.aml file.
> >   
> 
> We’ll address all the above when we add the CDAT.aml file support 
> which may only pass a subset of structures. 
> 
> >> +
> >> +    /* Set the DSLBIS entry, ent = 2 */
> >> +    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
> >> +    cxl_cstate->dslbis.header.reserved = 0;
> >> +    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
> >> +    cxl_cstate->dslbis.handle = 0;
> >> +    cxl_cstate->dslbis.flags = 0;
> >> +    cxl_cstate->dslbis.data_type = 0;
> >> +    cxl_cstate->dslbis.reserved2 = 0;
> >> +    cxl_cstate->dslbis.entry_base_unit = 0;
> >> +    cxl_cstate->dslbis.entry[0] = 0;
> >> +    cxl_cstate->dslbis.entry[1] = 0;
> >> +    cxl_cstate->dslbis.entry[2] = 0;
> >> +    cxl_cstate->dslbis.reserved3 = 0;
> >> +
> >> +    /* Set the DSMSCIS entry, ent = 3 */
> >> +    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
> >> +    cxl_cstate->dsmscis.header.reserved = 0;
> >> +    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
> >> +    cxl_cstate->dsmscis.DSMASH_handle = 0;
> >> +    cxl_cstate->dsmscis.reserved2[0] = 0;
> >> +    cxl_cstate->dsmscis.reserved2[1] = 0;
> >> +    cxl_cstate->dsmscis.reserved2[2] = 0;
> >> +    cxl_cstate->dsmscis.memory_side_cache_size = 0;
> >> +    cxl_cstate->dsmscis.cache_attributes = 0;
> >> +
> >> +    /* Set the DSIS entry, ent = 4 */
> >> +    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
> >> +    cxl_cstate->dsis.header.reserved = 0;
> >> +    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
> >> +    cxl_cstate->dsis.flags = 0;
> >> +    cxl_cstate->dsis.handle = 0;
> >> +    cxl_cstate->dsis.reserved2 = 0;
> >> +
> >> +    /* Set the DSEMTS entry, ent = 5 */
> >> +    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
> >> +    cxl_cstate->dsemts.header.reserved = 0;
> >> +    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
> >> +    cxl_cstate->dsemts.DSMAS_handle = 0;
> >> +    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
> >> +    cxl_cstate->dsemts.reserved2 = 0;
> >> +    cxl_cstate->dsemts.DPA_offset = 0;
> >> +    cxl_cstate->dsemts.DPA_length = 0;
> >> +
> >> +    /* Set the SSLBIS entry, ent = 6 */
> >> +    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
> >> +    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
> >> +    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
> >> +    cxl_cstate->sslbis.sslbis_h.data_type = 0;
> >> +    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
> >> +    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
> >> +    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
> >> +    /* Set the SSLBE entry */
> >> +    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
> >> +    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
> >> +    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
> >> +    cxl_cstate->sslbis.sslbe[0].reserved = 0;
> >> +    /* Set the SSLBE entry */
> >> +    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
> >> +    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
> >> +    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
> >> +    cxl_cstate->sslbis.sslbe[1].reserved = 0;
> >> +    /* Set the SSLBE entry */
> >> +    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
> >> +    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
> >> +    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
> >> +    cxl_cstate->sslbis.sslbe[2].reserved = 0;
> >> +
> >> +    /* Set CDAT header, ent = 0 */
> >> +    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
> >> +    cxl_cstate->cdat_header.reserved[0] = 0;
> >> +    cxl_cstate->cdat_header.reserved[1] = 0;
> >> +    cxl_cstate->cdat_header.reserved[2] = 0;
> >> +    cxl_cstate->cdat_header.reserved[3] = 0;
> >> +    cxl_cstate->cdat_header.reserved[4] = 0;
> >> +    cxl_cstate->cdat_header.reserved[5] = 0;
> >> +    cxl_cstate->cdat_header.sequence = 0;
> >> +
> >> +    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
> >> +        /* Add length of each CDAT struct to total length */
> >> +        len = cxl_cstate->cdat_ent[i].length;
> >> +        cxl_cstate->cdat_header.length += len;
> >> +
> >> +        /* Calculate checksum of each CDAT struct */
> >> +        for (j = 0; j < len; j++) {
> >> +            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
> >> +        }
> >> +    }
> >> +    cxl_cstate->cdat_header.checksum = ~sum + 1;
> >> +}
> >> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
> >> index d091e64..86c762f 100644
> >> --- a/hw/mem/cxl_type3.c
> >> +++ b/hw/mem/cxl_type3.c
> >> @@ -13,6 +13,150 @@
> >> #include "qemu/rcu.h"
> >> #include "sysemu/hostmem.h"
> >> #include "hw/cxl/cxl.h"
> >> +#include "hw/pci/msi.h"
> >> +#include "hw/pci/msix.h"
> >> +
> >> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap)  
> > local to file, static and remove from header.  
> >> +{
> >> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> >> +    uint32_t req;
> >> +    uint32_t byte_cnt = 0;
> >> +
> >> +    DOE_DBG(">> %s\n",  __func__);
> >> +
> >> +    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
> >> +        ->req_code;
> >> +    switch (req) {
> >> +    case CXL_COMP_MODE_CAP:
> >> +        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);  
> > 
> > Use a local variable to cap_rsp or assign it via a structure here.
> > Basically get rid of the repitition of
> > cxl_cstate->doe_resp.cap_rsp
> > in the interests of readability.
> > 
> >   
> >> +        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
> >> +        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
> >> +        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
> >> +        cxl_cstate->doe_resp.cap_rsp.header.length =
> >> +            dwsizeof(struct cxl_compliance_mode_cap_rsp);
> >> +        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
> >> +        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
> >> +        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
> >> +        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
> >> +        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
> >> +        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
> >> +        break;
> >> +    case CXL_COMP_MODE_STATUS:
> >> +        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
> >> +        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
> >> +        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
> >> +        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
> >> +        cxl_cstate->doe_resp.status_rsp.header.length =
> >> +            dwsizeof(struct cxl_compliance_mode_status_rsp);
> >> +        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
> >> +        cxl_cstate->doe_resp.status_rsp.version = 0x1;
> >> +        cxl_cstate->doe_resp.status_rsp.length = 0x14;
> >> +        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
> >> +        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
> >> +        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
> >> +        break;
> >> +    default:  
> > 
> > I guess the intent at somepoint is to actually hook some of these up?
> >   
> >> +        break;
> >> +    }
> >> +
> >> +    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
> >> +    DOE_DBG("<< %s\n",  __func__);
> >> +    return byte_cnt;
> >> +}
> >> +
> >> +
> >> +bool cxl_doe_compliance_rsp(DOECap *doe_cap)  
> > 
> > Currently this is local to this file, so drop it form the header and
> > mark it static.  
> >   
> >> +{
> >> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> >> +    uint32_t byte_cnt;
> >> +    uint32_t dw_cnt;
> >> +
> >> +    DOE_DBG(">> %s\n",  __func__);
> >> +
> >> +    byte_cnt = cxl_doe_compliance_init(doe_cap);
> >> +    dw_cnt = byte_cnt / 4;
> >> +    memcpy(doe_cap->read_mbox,
> >> +           cxl_cstate->doe_resp.data_byte, byte_cnt);
> >> +
> >> +    doe_cap->read_mbox_len += dw_cnt;
> >> +
> >> +    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
> >> +            doe_cap->read_mbox_len,
> >> +            *(doe_cap->read_mbox + dw_cnt - 1));
> >> +
> >> +    DOE_DBG("<< %s\n",  __func__);
> >> +
> >> +    return DOE_SUCCESS;
> >> +}
> >> +
> >> +bool cxl_doe_cdat_rsp(DOECap *doe_cap)  
> > Local to this file I think so drop it from header and mark it static.
> >   
> >> +{
> >> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
> >> +    uint16_t ent;
> >> +    void *base;
> >> +    uint32_t len;
> >> +    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
> >> +
> >> +    ent = req->entry_handle;
> >> +    base = cxl_cstate->cdat_ent[ent].base;
> >> +    len = cxl_cstate->cdat_ent[ent].length;
> >> +
> >> +    struct cxl_cdat_rsp rsp = {
> >> +        .header = {
> >> +            .vendor_id = CXL_VENDOR_ID,
> >> +            .doe_type = CXL_DOE_TABLE_ACCESS,
> >> +            .reserved = 0x0,
> >> +            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
> >> +        },
> >> +        .req_code = CXL_DOE_TAB_RSP,
> >> +        .table_type = CXL_DOE_TAB_TYPE_CDAT,
> >> +        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
> >> +    };
> >> +
> >> +    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
> >> +    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
> >> +
> >> +    doe_cap->read_mbox_len += rsp.header.length;
> >> +    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
> >> +
> >> +    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
> >> +        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
> >> +    }
> >> +
> >> +    return DOE_SUCCESS;
> >> +}
> >> +
> >> +static uint32_t ct3d_config_read(PCIDevice *pci_dev,
> >> +                            uint32_t addr, int size)
> >> +{
> >> +    CXLType3Dev *ct3d = CT3(pci_dev);
> >> +    PCIEDOE *doe = &ct3d->doe;
> >> +    DOECap *doe_cap;
> >> +
> >> +    doe_cap = pcie_doe_covers_addr(doe, addr);
> >> +    if (doe_cap) {
> >> +        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
> >> +        return pcie_doe_read_config(doe_cap, addr, size);
> >> +    } else {
> >> +        return pci_default_read_config(pci_dev, addr, size);
> >> +    }
> >> +}
> >> +
> >> +static void ct3d_config_write(PCIDevice *pci_dev,
> >> +                            uint32_t addr, uint32_t val, int size)
> >> +{
> >> +    CXLType3Dev *ct3d = CT3(pci_dev);
> >> +    PCIEDOE *doe = &ct3d->doe;
> >> +    DOECap *doe_cap;
> >> +
> >> +    doe_cap = pcie_doe_covers_addr(doe, addr);
> >> +    if (doe_cap) {
> >> +        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
> >> +        pcie_doe_write_config(doe_cap, addr, val, size);
> >> +    } else {
> >> +        pci_default_write_config(pci_dev, addr, val, size);
> >> +    }
> >> +}
> >> 
> >> static void build_dvsecs(CXLType3Dev *ct3d)
> >> {
> >> @@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
> >>     ComponentRegisters *regs = &cxl_cstate->crb;
> >>     MemoryRegion *mr = &regs->component_registers;
> >>     uint8_t *pci_conf = pci_dev->config;
> >> +    unsigned short msix_num = 2;
> >> +    int rc;
> >> +    int i;
> >> 
> >>     if (!ct3d->cxl_dstate.pmem) {
> >>         cxl_setup_memory(ct3d, errp);
> >> @@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
> >>                      PCI_BASE_ADDRESS_SPACE_MEMORY |
> >>                          PCI_BASE_ADDRESS_MEM_TYPE_64,
> >>                      &ct3d->cxl_dstate.device_registers);
> >> +
> >> +    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
> >> +    for (i = 0; i < msix_num; i++) {
> >> +        msix_vector_use(pci_dev, i);
> >> +    }
> >> +
> >> +    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */  
> > 
> > Tidy this up or parameterize it.
> >   
> >> +
> >> +    pcie_doe_init(pci_dev, &ct3d->doe);
> >> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);  
> > 
> > check rc here.
> >   
> >> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
> >> +    if (rc) {
> >> +        error_setg(errp, "fail to add DOE cap");
> >> +        return;
> >> +    }
> >> +
> >> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
> >> +                               cxl_doe_compliance_rsp);
> >> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
> >> +                               cxl_doe_cdat_rsp);
> >> +
> >> +    cxl_doe_cdat_init(cxl_cstate);
> >> }
> >> 
> >> static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
> >> @@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
> >>     DeviceClass *dc = DEVICE_CLASS(oc);
> >>     PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
> >>     MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
> >> +
> >> +    pc->config_write = ct3d_config_write;
> >> +    pc->config_read = ct3d_config_read;
> >>     CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
> >> 
> >>     pc->realize = ct3_realize;
> >> diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
> >> new file mode 100644
> >> index 0000000..fbbd494
> >> --- /dev/null
> >> +++ b/include/hw/cxl/cxl_cdat.h
> >> @@ -0,0 +1,120 @@
> >> +#include "hw/cxl/cxl_pci.h"
> >> +
> >> +
> >> +enum {
> >> +    CXL_DOE_COMPLIANCE             = 0,
> >> +    CXL_DOE_TABLE_ACCESS           = 2,
> >> +    CXL_DOE_MAX_PROTOCOL
> >> +};
> >> +
> >> +#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
> >> +#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
> >> +
> >> +/*
> >> + * DOE CDAT Table Protocol (CXL Spec)
> >> + */
> >> +#define CXL_DOE_TAB_REQ 0
> >> +#define CXL_DOE_TAB_RSP 0
> >> +#define CXL_DOE_TAB_TYPE_CDAT 0
> >> +#define CXL_DOE_TAB_ENT_MAX 0xFFFF
> >> +
> >> +/* Read Entry Request, 8.1.11.1 Table 134 */
> >> +struct cxl_cdat {
> >> +    DOEHeader header;
> >> +    uint8_t req_code;
> >> +    uint8_t table_type;
> >> +    uint16_t entry_handle;
> >> +} QEMU_PACKED;
> >> +
> >> +/* Read Entry Response, 8.1.11.1 Table 135 */
> >> +#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
> >> +  
> > ... Note I'm just snipping out these big defines as I check them
> > against the spec :) Hence the gap.
> > ...
> >   
> >> +struct cdat_dsmscis {
> >> +    struct cdat_sub_header header;
> >> +    uint8_t DSMASH_handle;  
> > 
> > DMSAS_handle;
> >   
> >> +    uint8_t reserved2[3];
> >> +    uint64_t memory_side_cache_size;
> >> +    uint32_t cache_attributes;
> >> +} QEMU_PACKED;
> >> +  
> > 
> >   
> >> +
> >> +struct cdat_sslbis_header {
> >> +    struct cdat_sub_header header;
> >> +    uint8_t data_type;
> >> +    uint8_t reserved2[3];
> >> +    uint64_t entry_base_unit;
> >> +} QEMU_PACKED;
> >> diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
> >> new file mode 100644
> >> index 0000000..ebbe488
> >> --- /dev/null
> >> +++ b/include/hw/cxl/cxl_compl.h
> >> @@ -0,0 +1,289 @@
> >> +/*
> >> + * CXL Compliance Mode Protocol  
> > 
> > Needs license etc I think
> >   
> >> + */  
> > 
> > A bunch of the responses below are all of this form, perhaps we
> > could just have one cxl_compliance_mode_generic_status_rsp ?
> > if you really want to then define the others perhaps use
> > #define to do it rather than lots of identical structures
> > each specified fully.  
> 
> 
> Got it.   Should have simplified it after cutting and pasting so many times.
> > 
> >   
> >> +struct cxl_compliance_mode_inject_mac_delay_rsp {
> >> +    DOEHeader header;
> >> +    uint8_t rsp_code;
> >> +    uint8_t version;
> >> +    uint8_t length;
> >> +    uint8_t status;
> >> +} QEMU_PACKED;
> >> +  
> > 
> > 
> >   
> >> +struct cxl_compliance_mode_ignore_bit_error {
> >> +    DOEHeader header;
> >> +    uint8_t req_code;
> >> +    uint8_t version;
> >> +    uint16_t reserved;
> >> +    uint8_t opcode;
> >> +} QEMU_PACKED;
> >> +  
> > This last one doesn't seem to line up with the CXL 2.0 spec.  
> 
> Good catch!
> 
> >   
> >> +struct cxl_compliance_mode_ignore_bit_error_rsp {
> >> +    DOEHeader header;
> >> +    uint8_t rsp_code;
> >> +    uint8_t version;
> >> +    uint8_t reserved[6];
> >> +} QEMU_PACKED;  
> > 
> >   
> >> diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
> >> index 762feb5..23923df 100644
> >> --- a/include/hw/cxl/cxl_component.h
> >> +++ b/include/hw/cxl/cxl_component.h
> >> @@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
> >> _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
> >>                "No space for registers");  
> > 
> > ...
> >   
> >> typedef struct component_registers {
> >>     /*
> >>      * Main memory region to be registered with QEMU core.
> >> @@ -160,6 +177,10 @@ typedef struct component_registers {
> >>     MemoryRegionOps *special_ops;
> >> } ComponentRegisters;
> >> 
> >> +typedef struct cdat_struct {
> >> +    void *base;
> >> +    uint32_t length;
> >> +} CDATStruct;
> >> /*
> >>  * A CXL component represents all entities in a CXL hierarchy. This includes,
> >>  * host bridges, root ports, upstream/downstream switch ports, and devices
> >> @@ -173,6 +194,104 @@ typedef struct cxl_component {
> >>             struct PCIDevice *pdev;
> >>         };
> >>     };
> >> +
> >> +    /*
> >> +     * SW write write mailbox and GO on last DW
> >> +     * device sets READY of offset DW in DO types to copy
> >> +     * to read mailbox register on subsequent cfg_read
> >> +     * of read mailbox register and then on cfg_write to
> >> +     * denote success read increments offset to next DW
> >> +     */
> >> +
> >> +    union doe_request_u {
> >> +        /* Compliance DOE Data Objects Type=0*/
> >> +        struct cxl_compliance_mode_cap
> >> +            mode_cap;  
> > 
> > I'd only add line breaks for the longer ones of these.
> >   
> >> +        struct cxl_compliance_mode_status
> >> +            mode_status;
> >> +        struct cxl_compliance_mode_halt
> >> +            mode_halt;
> >> +        struct cxl_compliance_mode_multiple_write_streaming
> >> +            multiple_write_streaming;
> >> +        struct cxl_compliance_mode_producer_consumer
> >> +            producer_consumer;
> >> +        struct cxl_compliance_mode_inject_bogus_writes
> >> +            inject_bogus_writes;
> >> +        struct cxl_compliance_mode_inject_poison
> >> +            inject_poison;
> >> +        struct cxl_compliance_mode_inject_crc
> >> +            inject_crc;
> >> +        struct cxl_compliance_mode_inject_flow_control
> >> +            inject_flow_control;
> >> +        struct cxl_compliance_mode_toggle_cache_flush
> >> +            toggle_cache_flush;
> >> +        struct cxl_compliance_mode_inject_mac_delay
> >> +            inject_mac_delay;
> >> +        struct cxl_compliance_mode_insert_unexp_mac
> >> +            insert_unexp_mac;
> >> +        struct cxl_compliance_mode_inject_viral
> >> +            inject_viral;
> >> +        struct cxl_compliance_mode_inject_almp
> >> +            inject_almp;
> >> +        struct cxl_compliance_mode_ignore_almp
> >> +            ignore_almp;
> >> +        struct cxl_compliance_mode_ignore_bit_error
> >> +            ignore_bit_error;
> >> +        char data_byte[128];
> >> +    } doe_request;
> >> +
> >> +    union doe_resp_u {
> >> +        /* Compliance DOE Data Objects Type=0*/
> >> +        struct cxl_compliance_mode_cap_rsp
> >> +            cap_rsp;
> >> +        struct cxl_compliance_mode_status_rsp
> >> +            status_rsp;
> >> +        struct cxl_compliance_mode_halt_rsp
> >> +            halt_rsp;
> >> +        struct cxl_compliance_mode_multiple_write_streaming_rsp
> >> +            multiple_write_streaming_rsp;
> >> +        struct cxl_compliance_mode_producer_consumer_rsp
> >> +            producer_consumer_rsp;
> >> +        struct cxl_compliance_mode_inject_bogus_writes_rsp
> >> +            inject_bogus_writes_rsp;
> >> +        struct cxl_compliance_mode_inject_poison_rsp
> >> +            inject_poison_rsp;
> >> +        struct cxl_compliance_mode_inject_crc_rsp
> >> +            inject_crc_rsp;
> >> +        struct cxl_compliance_mode_inject_flow_control_rsp
> >> +            inject_flow_control_rsp;
> >> +        struct cxl_compliance_mode_toggle_cache_flush_rsp
> >> +            toggle_cache_flush_rsp;
> >> +        struct cxl_compliance_mode_inject_mac_delay_rsp
> >> +            inject_mac_delay_rsp;
> >> +        struct cxl_compliance_mode_insert_unexp_mac_rsp
> >> +            insert_unexp_mac_rsp;
> >> +        struct cxl_compliance_mode_inject_viral
> >> +            inject_viral_rsp;
> >> +        struct cxl_compliance_mode_inject_almp_rsp
> >> +            inject_almp_rsp;
> >> +        struct cxl_compliance_mode_ignore_almp_rsp
> >> +            ignore_almp_rsp;
> >> +        struct cxl_compliance_mode_ignore_bit_error_rsp
> >> +            ignore_bit_error_rsp;
> >> +        char data_byte[520 * 4];
> >> +        uint32_t data_dword[520];
> >> +    } doe_resp;
> >> +
> >> +    /* Table entry types */  
> > Hmm. Not sure all CXL components will have CDAT.  Root ports for
> > example?
> >   
> >> +    struct cdat_table_header cdat_header;
> >> +    struct cdat_dsmas dsmas;
> >> +    struct cdat_dslbis dslbis;  
> > 
> > As I said in v1, some of these will need to be multiples so this
> > flat structure just doesn't work.
> >   
> >> +    struct cdat_dsmscis dsmscis;
> >> +    struct cdat_dsis dsis;
> >> +    struct cdat_dsemts dsemts;
> >> +    struct {
> >> +        struct cdat_sslbis_header sslbis_h;
> >> +        struct cdat_sslbe sslbe[3];  
> > 
> > I'm curious.  Why 3?  This is between pairs of ports so 1USP 2DSP switch?
> >   
> >> +    } sslbis;  
> > 
> >   
> >> +
> >> +    CDATStruct *cdat_ent;
> >> +    int cdat_ent_len;
> >> } CXLComponentState;
> >> 
> >> void cxl_component_register_block_init(Object *obj,
> >> @@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
> >> void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
> >>                                 uint16_t type, uint8_t rev, uint8_t *body);
> >> 
> >> +void cxl_component_create_doe(CXLComponentState *cxl_cstate,
> >> +                              uint16_t offset, unsigned vec);  
> > 
> > Doesn't seem to exist.
> > 
> > Some of the following are local to one file so shouldn't be here iether.
> >   
> >> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
> >> +bool cxl_doe_compliance_rsp(DOECap *doe_cap);
> >> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
> >> +bool cxl_doe_cdat_rsp(DOECap *doe_cap);
> >> +uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);  
> > 
> > Doesn't seem to exist.
> >   
> >> #endif
> >> diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
> >> index c608ced..08bf646 100644
> >> --- a/include/hw/cxl/cxl_device.h
> >> +++ b/include/hw/cxl/cxl_device.h
> >> @@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
> >>     /* State */
> >>     CXLComponentState cxl_cstate;
> >>     CXLDeviceState cxl_dstate;
> >> +
> >> +    /* DOE */
> >> +    PCIEDOE doe;
> >> } CXLType3Dev;
> >> 
> >> #ifndef TYPE_CXL_TYPE3_DEV
> >> diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
> >> index 9ec28c9..5cab197 100644
> >> --- a/include/hw/cxl/cxl_pci.h
> >> +++ b/include/hw/cxl/cxl_pci.h
> >> @@ -12,6 +12,8 @@
> >> 
> >> #include "hw/pci/pci.h"
> >> #include "hw/pci/pcie.h"
> >> +#include "hw/cxl/cxl_cdat.h"
> >> +#include "hw/cxl/cxl_compl.h"
> >> 
> >> #define CXL_VENDOR_ID 0x1e98
> >> 
> >> @@ -54,6 +56,8 @@ struct dvsec_header {
> >> _Static_assert(sizeof(struct dvsec_header) == 10,
> >>                "dvsec header size incorrect");
> >> 
> >> +/* CXL 2.0 - 8.1.11 */
> >> +  
> > 
> > Clean this out next time - doesn't belong in this patch.
> >   
> >> /*
> >>  * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
> >>  * implement others.  
> >   
> 

Re: [RFC v2 2/2] Basic CXL DOE for CDAT and Compliance Mode

[Chris Browy]

> On Feb 18, 2021, at 2:15 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
> 
> On Fri, 12 Feb 2021 17:26:50 -0500
> Chris Browy <cbrowy@avery-design.com> wrote:
> 
>>> On Feb 12, 2021, at 12:23 PM, Jonathan Cameron <jonathan.cameron@huawei.com> wrote:
>>> 
>>> On Tue, 9 Feb 2021 15:36:03 -0500
>>> Chris Browy <cbrowy@avery-design.com> wrote:
>>> 
>>> Split this into two patches for v3.  CDAT in one, compliance mode in the other.
>>> 
>> 
>> Compliance mode is an optional feature.  We’ll split it out.
>> 
>>> I'd also move the actual elements out into the cxl components so that we
>>> can register only what makes sense for a given device.   My guess
>>> is that for now that will be static const anyway.
>>> 
>>> Coming together fine. Hopefully I'll start poking at the linux side of things
>>> next week.  First job being simply providing a file to allow us to dump
>>> the whole CDAT table.  Let me know if you get this loading an .aml file
>>> in the meantime as that'll make it easier to test (if not I'll hack it
>>> on top of these patches)  
>> 
>> We can get the .aml loading by Thurs next week.  Holiday next few days for 
>> some of our folks.
>> 
>>> 
>>> If needed I'll add it to iASL as well (may well be already in hand!)
> 
> There is a potential problem doing this.  CDAT doesn't have the table
> type ID that an ACPI table would have.  That means raw CDAT tables
> are not identifiable and I think this makes it hard to use iASL with them
> without changing it's general means of functioning.
> 
> We can probably do something with an extra parameter, but this lack of
> identifier is going to make it harder to persuade people that it's sensible to
> including CDAT in iASL.

This would be worth requesting the responsible ACPI or UEFI working group 
of UEFI.org to weight in on the original intent since this must have been considered
despite not being addressed in the specification.

The spec is clear:

	Note: The data structures defined in this document are NOT ACPI tables.

I can’t find the author or working group designator in the spec although it is
	Copyright 2020 Unified EFI, Inc. All Rights Reserved.

> 
>>> 
>>> I think my version of this stuff did a useful job in improving my understanding
>>> of what we were trying to do, but that done I'm assuming we'll just abandon it
>>> as the disposable prototype it was :)
>>> 
>> 
>> Thanks for focusing in on the area and uncovering problems with both our versions!
>> 
>> Still lots of pieces need to come together and get working to be able to fully enumerate 
>> and configure the device!
>> 
>>> Jonathan
>>> 
>>> 
>>>> ---
>>>> hw/cxl/cxl-component-utils.c   | 132 +++++++++++++++++++
>>>> hw/mem/cxl_type3.c             | 172 ++++++++++++++++++++++++
>>>> include/hw/cxl/cxl_cdat.h      | 120 +++++++++++++++++
>>>> include/hw/cxl/cxl_compl.h     | 289 +++++++++++++++++++++++++++++++++++++++++
>>>> include/hw/cxl/cxl_component.h | 126 ++++++++++++++++++
>>>> include/hw/cxl/cxl_device.h    |   3 +
>>>> include/hw/cxl/cxl_pci.h       |   4 +
>>>> 7 files changed, 846 insertions(+)
>>>> create mode 100644 include/hw/cxl/cxl_cdat.h
>>>> create mode 100644 include/hw/cxl/cxl_compl.h
>>>> 
>>>> diff --git a/hw/cxl/cxl-component-utils.c b/hw/cxl/cxl-component-utils.c
>>>> index e1bcee5..fc6c538 100644
>>>> --- a/hw/cxl/cxl-component-utils.c
>>>> +++ b/hw/cxl/cxl-component-utils.c
>>>> @@ -195,3 +195,135 @@ void cxl_component_create_dvsec(CXLComponentState *cxl, uint16_t length,
>>>>    range_init_nofail(&cxl->dvsecs[type], cxl->dvsec_offset, length);
>>>>    cxl->dvsec_offset += length;
>>>> }
>>>> +
>>>> +/* Return the sum of bytes */
>>>> +static void cdat_ent_init(CDATStruct *cs, void *base, uint32_t len)
>>>> +{
>>>> +    cs->base = base;
>>>> +    cs->length = len;
>>>> +}
>>>> +
>>>> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate)
>>>> +{
>>>> +    uint8_t sum = 0;
>>>> +    uint32_t len = 0;
>>>> +    int i, j;
>>>> +
>>>> +    cxl_cstate->cdat_ent_len = 7;
>>>> +    cxl_cstate->cdat_ent =
>>>> +        g_malloc0(sizeof(CDATStruct) * cxl_cstate->cdat_ent_len);
>>>> +
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[0],
>>>> +                  &cxl_cstate->cdat_header, sizeof(cxl_cstate->cdat_header));
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[1],
>>>> +                  &cxl_cstate->dsmas, sizeof(cxl_cstate->dsmas));
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[2],
>>>> +                  &cxl_cstate->dslbis, sizeof(cxl_cstate->dslbis));
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[3],
>>>> +                  &cxl_cstate->dsmscis, sizeof(cxl_cstate->dsmscis));
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[4],
>>>> +                  &cxl_cstate->dsis, sizeof(cxl_cstate->dsis));
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[5],
>>>> +                  &cxl_cstate->dsemts, sizeof(cxl_cstate->dsemts));
>>>> +    cdat_ent_init(&cxl_cstate->cdat_ent[6],
>>>> +                  &cxl_cstate->sslbis, sizeof(cxl_cstate->sslbis));
>>>> +
>>>> +    /* Set the DSMAS entry, ent = 1 */
>>>> +    cxl_cstate->dsmas.header.type = CDAT_TYPE_DSMAS;
>>>> +    cxl_cstate->dsmas.header.reserved = 0x0;
>>>> +    cxl_cstate->dsmas.header.length = sizeof(cxl_cstate->dsmas);
>>>> +    cxl_cstate->dsmas.DSMADhandle = 0x0;
>>>> +    cxl_cstate->dsmas.flags = 0x0;
>>>> +    cxl_cstate->dsmas.reserved2 = 0x0;
>>>> +    cxl_cstate->dsmas.DPA_base = 0x0;
>>>> +    cxl_cstate->dsmas.DPA_length = 0x40000;  
>>> 
>>> Look to move the instances of these down into the memory device and expose
>>> cdat_ent_init() to there.
>>> 
>>> That way, we can add whatever elements make sense for each type
>>> of component.
>>> 
>> 
>>> Also have a cdat_ents_finalize() or similar to call at the end
>>> which calculates overall length + checksum.  
>> 
>> OK
>> 
>>> 
>>> Should also be easy enough to add a simple bit of code to call
>>> cdat_ent_init() for each element of a passed in CDAT.aml file.
>>> 
>> 
>> We’ll address all the above when we add the CDAT.aml file support 
>> which may only pass a subset of structures. 
>> 
>>>> +
>>>> +    /* Set the DSLBIS entry, ent = 2 */
>>>> +    cxl_cstate->dslbis.header.type = CDAT_TYPE_DSLBIS;
>>>> +    cxl_cstate->dslbis.header.reserved = 0;
>>>> +    cxl_cstate->dslbis.header.length = sizeof(cxl_cstate->dslbis);
>>>> +    cxl_cstate->dslbis.handle = 0;
>>>> +    cxl_cstate->dslbis.flags = 0;
>>>> +    cxl_cstate->dslbis.data_type = 0;
>>>> +    cxl_cstate->dslbis.reserved2 = 0;
>>>> +    cxl_cstate->dslbis.entry_base_unit = 0;
>>>> +    cxl_cstate->dslbis.entry[0] = 0;
>>>> +    cxl_cstate->dslbis.entry[1] = 0;
>>>> +    cxl_cstate->dslbis.entry[2] = 0;
>>>> +    cxl_cstate->dslbis.reserved3 = 0;
>>>> +
>>>> +    /* Set the DSMSCIS entry, ent = 3 */
>>>> +    cxl_cstate->dsmscis.header.type = CDAT_TYPE_DSMSCIS;
>>>> +    cxl_cstate->dsmscis.header.reserved = 0;
>>>> +    cxl_cstate->dsmscis.header.length = sizeof(cxl_cstate->dsmscis);
>>>> +    cxl_cstate->dsmscis.DSMASH_handle = 0;
>>>> +    cxl_cstate->dsmscis.reserved2[0] = 0;
>>>> +    cxl_cstate->dsmscis.reserved2[1] = 0;
>>>> +    cxl_cstate->dsmscis.reserved2[2] = 0;
>>>> +    cxl_cstate->dsmscis.memory_side_cache_size = 0;
>>>> +    cxl_cstate->dsmscis.cache_attributes = 0;
>>>> +
>>>> +    /* Set the DSIS entry, ent = 4 */
>>>> +    cxl_cstate->dsis.header.type = CDAT_TYPE_DSIS;
>>>> +    cxl_cstate->dsis.header.reserved = 0;
>>>> +    cxl_cstate->dsis.header.length = sizeof(cxl_cstate->dsis);
>>>> +    cxl_cstate->dsis.flags = 0;
>>>> +    cxl_cstate->dsis.handle = 0;
>>>> +    cxl_cstate->dsis.reserved2 = 0;
>>>> +
>>>> +    /* Set the DSEMTS entry, ent = 5 */
>>>> +    cxl_cstate->dsemts.header.type = CDAT_TYPE_DSEMTS;
>>>> +    cxl_cstate->dsemts.header.reserved = 0;
>>>> +    cxl_cstate->dsemts.header.length = sizeof(cxl_cstate->dsemts);
>>>> +    cxl_cstate->dsemts.DSMAS_handle = 0;
>>>> +    cxl_cstate->dsemts.EFI_memory_type_attr = 0;
>>>> +    cxl_cstate->dsemts.reserved2 = 0;
>>>> +    cxl_cstate->dsemts.DPA_offset = 0;
>>>> +    cxl_cstate->dsemts.DPA_length = 0;
>>>> +
>>>> +    /* Set the SSLBIS entry, ent = 6 */
>>>> +    cxl_cstate->sslbis.sslbis_h.header.type = CDAT_TYPE_SSLBIS;
>>>> +    cxl_cstate->sslbis.sslbis_h.header.reserved = 0;
>>>> +    cxl_cstate->sslbis.sslbis_h.header.length = sizeof(cxl_cstate->sslbis);
>>>> +    cxl_cstate->sslbis.sslbis_h.data_type = 0;
>>>> +    cxl_cstate->sslbis.sslbis_h.reserved2[0] = 0;
>>>> +    cxl_cstate->sslbis.sslbis_h.reserved2[1] = 0;
>>>> +    cxl_cstate->sslbis.sslbis_h.reserved2[2] = 0;
>>>> +    /* Set the SSLBE entry */
>>>> +    cxl_cstate->sslbis.sslbe[0].port_x_id = 0;
>>>> +    cxl_cstate->sslbis.sslbe[0].port_y_id = 0;
>>>> +    cxl_cstate->sslbis.sslbe[0].latency_bandwidth = 0;
>>>> +    cxl_cstate->sslbis.sslbe[0].reserved = 0;
>>>> +    /* Set the SSLBE entry */
>>>> +    cxl_cstate->sslbis.sslbe[1].port_x_id = 1;
>>>> +    cxl_cstate->sslbis.sslbe[1].port_y_id = 1;
>>>> +    cxl_cstate->sslbis.sslbe[1].latency_bandwidth = 0;
>>>> +    cxl_cstate->sslbis.sslbe[1].reserved = 0;
>>>> +    /* Set the SSLBE entry */
>>>> +    cxl_cstate->sslbis.sslbe[2].port_x_id = 2;
>>>> +    cxl_cstate->sslbis.sslbe[2].port_y_id = 2;
>>>> +    cxl_cstate->sslbis.sslbe[2].latency_bandwidth = 0;
>>>> +    cxl_cstate->sslbis.sslbe[2].reserved = 0;
>>>> +
>>>> +    /* Set CDAT header, ent = 0 */
>>>> +    cxl_cstate->cdat_header.revision = CXL_CDAT_REV;
>>>> +    cxl_cstate->cdat_header.reserved[0] = 0;
>>>> +    cxl_cstate->cdat_header.reserved[1] = 0;
>>>> +    cxl_cstate->cdat_header.reserved[2] = 0;
>>>> +    cxl_cstate->cdat_header.reserved[3] = 0;
>>>> +    cxl_cstate->cdat_header.reserved[4] = 0;
>>>> +    cxl_cstate->cdat_header.reserved[5] = 0;
>>>> +    cxl_cstate->cdat_header.sequence = 0;
>>>> +
>>>> +    for (i = cxl_cstate->cdat_ent_len - 1; i >= 0; i--) {
>>>> +        /* Add length of each CDAT struct to total length */
>>>> +        len = cxl_cstate->cdat_ent[i].length;
>>>> +        cxl_cstate->cdat_header.length += len;
>>>> +
>>>> +        /* Calculate checksum of each CDAT struct */
>>>> +        for (j = 0; j < len; j++) {
>>>> +            sum += *(uint8_t *)(cxl_cstate->cdat_ent[i].base + j);
>>>> +        }
>>>> +    }
>>>> +    cxl_cstate->cdat_header.checksum = ~sum + 1;
>>>> +}
>>>> diff --git a/hw/mem/cxl_type3.c b/hw/mem/cxl_type3.c
>>>> index d091e64..86c762f 100644
>>>> --- a/hw/mem/cxl_type3.c
>>>> +++ b/hw/mem/cxl_type3.c
>>>> @@ -13,6 +13,150 @@
>>>> #include "qemu/rcu.h"
>>>> #include "sysemu/hostmem.h"
>>>> #include "hw/cxl/cxl.h"
>>>> +#include "hw/pci/msi.h"
>>>> +#include "hw/pci/msix.h"
>>>> +
>>>> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap)  
>>> local to file, static and remove from header.  
>>>> +{
>>>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>>>> +    uint32_t req;
>>>> +    uint32_t byte_cnt = 0;
>>>> +
>>>> +    DOE_DBG(">> %s\n",  __func__);
>>>> +
>>>> +    req = ((struct cxl_compliance_mode_cap *)pcie_doe_get_req(doe_cap))
>>>> +        ->req_code;
>>>> +    switch (req) {
>>>> +    case CXL_COMP_MODE_CAP:
>>>> +        byte_cnt = sizeof(struct cxl_compliance_mode_cap_rsp);  
>>> 
>>> Use a local variable to cap_rsp or assign it via a structure here.
>>> Basically get rid of the repitition of
>>> cxl_cstate->doe_resp.cap_rsp
>>> in the interests of readability.
>>> 
>>> 
>>>> +        cxl_cstate->doe_resp.cap_rsp.header.vendor_id = CXL_VENDOR_ID;
>>>> +        cxl_cstate->doe_resp.cap_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
>>>> +        cxl_cstate->doe_resp.cap_rsp.header.reserved = 0x0;
>>>> +        cxl_cstate->doe_resp.cap_rsp.header.length =
>>>> +            dwsizeof(struct cxl_compliance_mode_cap_rsp);
>>>> +        cxl_cstate->doe_resp.cap_rsp.rsp_code = 0x0;
>>>> +        cxl_cstate->doe_resp.cap_rsp.version = 0x1;
>>>> +        cxl_cstate->doe_resp.cap_rsp.length = 0x1c;
>>>> +        cxl_cstate->doe_resp.cap_rsp.status = 0x0;
>>>> +        cxl_cstate->doe_resp.cap_rsp.available_cap_bitmask = 0x3;
>>>> +        cxl_cstate->doe_resp.cap_rsp.enabled_cap_bitmask = 0x3;
>>>> +        break;
>>>> +    case CXL_COMP_MODE_STATUS:
>>>> +        byte_cnt = sizeof(struct cxl_compliance_mode_status_rsp);
>>>> +        cxl_cstate->doe_resp.status_rsp.header.vendor_id = CXL_VENDOR_ID;
>>>> +        cxl_cstate->doe_resp.status_rsp.header.doe_type = CXL_DOE_COMPLIANCE;
>>>> +        cxl_cstate->doe_resp.status_rsp.header.reserved = 0x0;
>>>> +        cxl_cstate->doe_resp.status_rsp.header.length =
>>>> +            dwsizeof(struct cxl_compliance_mode_status_rsp);
>>>> +        cxl_cstate->doe_resp.status_rsp.rsp_code = 0x1;
>>>> +        cxl_cstate->doe_resp.status_rsp.version = 0x1;
>>>> +        cxl_cstate->doe_resp.status_rsp.length = 0x14;
>>>> +        cxl_cstate->doe_resp.status_rsp.cap_bitfield = 0x3;
>>>> +        cxl_cstate->doe_resp.status_rsp.cache_size = 0;
>>>> +        cxl_cstate->doe_resp.status_rsp.cache_size_units = 0;
>>>> +        break;
>>>> +    default:  
>>> 
>>> I guess the intent at somepoint is to actually hook some of these up?
>>> 
>>>> +        break;
>>>> +    }
>>>> +
>>>> +    DOE_DBG("  REQ=%x, RSP BYTE_CNT=%d\n", req, byte_cnt);
>>>> +    DOE_DBG("<< %s\n",  __func__);
>>>> +    return byte_cnt;
>>>> +}
>>>> +
>>>> +
>>>> +bool cxl_doe_compliance_rsp(DOECap *doe_cap)  
>>> 
>>> Currently this is local to this file, so drop it form the header and
>>> mark it static.  
>>> 
>>>> +{
>>>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>>>> +    uint32_t byte_cnt;
>>>> +    uint32_t dw_cnt;
>>>> +
>>>> +    DOE_DBG(">> %s\n",  __func__);
>>>> +
>>>> +    byte_cnt = cxl_doe_compliance_init(doe_cap);
>>>> +    dw_cnt = byte_cnt / 4;
>>>> +    memcpy(doe_cap->read_mbox,
>>>> +           cxl_cstate->doe_resp.data_byte, byte_cnt);
>>>> +
>>>> +    doe_cap->read_mbox_len += dw_cnt;
>>>> +
>>>> +    DOE_DBG("  LEN=%x, RD MBOX[%d]=%x\n", dw_cnt - 1,
>>>> +            doe_cap->read_mbox_len,
>>>> +            *(doe_cap->read_mbox + dw_cnt - 1));
>>>> +
>>>> +    DOE_DBG("<< %s\n",  __func__);
>>>> +
>>>> +    return DOE_SUCCESS;
>>>> +}
>>>> +
>>>> +bool cxl_doe_cdat_rsp(DOECap *doe_cap)  
>>> Local to this file I think so drop it from header and mark it static.
>>> 
>>>> +{
>>>> +    CXLComponentState *cxl_cstate = &CT3(doe_cap->doe->pdev)->cxl_cstate;
>>>> +    uint16_t ent;
>>>> +    void *base;
>>>> +    uint32_t len;
>>>> +    struct cxl_cdat *req = pcie_doe_get_req(doe_cap);
>>>> +
>>>> +    ent = req->entry_handle;
>>>> +    base = cxl_cstate->cdat_ent[ent].base;
>>>> +    len = cxl_cstate->cdat_ent[ent].length;
>>>> +
>>>> +    struct cxl_cdat_rsp rsp = {
>>>> +        .header = {
>>>> +            .vendor_id = CXL_VENDOR_ID,
>>>> +            .doe_type = CXL_DOE_TABLE_ACCESS,
>>>> +            .reserved = 0x0,
>>>> +            .length = (sizeof(struct cxl_cdat_rsp) + len) / 4,
>>>> +        },
>>>> +        .req_code = CXL_DOE_TAB_RSP,
>>>> +        .table_type = CXL_DOE_TAB_TYPE_CDAT,
>>>> +        .entry_handle = (++ent < cxl_cstate->cdat_ent_len) ? ent : CXL_DOE_TAB_ENT_MAX,
>>>> +    };
>>>> +
>>>> +    memcpy(doe_cap->read_mbox, &rsp, sizeof(rsp));
>>>> +    memcpy(doe_cap->read_mbox + sizeof(rsp) / 4, base, len);
>>>> +
>>>> +    doe_cap->read_mbox_len += rsp.header.length;
>>>> +    DOE_DBG("  read_mbox_len=%x\n", doe_cap->read_mbox_len);
>>>> +
>>>> +    for (int i = 0; i < doe_cap->read_mbox_len; i++) {
>>>> +        DOE_DBG("  RD MBOX[%d]=%08x\n", i, doe_cap->read_mbox[i]);
>>>> +    }
>>>> +
>>>> +    return DOE_SUCCESS;
>>>> +}
>>>> +
>>>> +static uint32_t ct3d_config_read(PCIDevice *pci_dev,
>>>> +                            uint32_t addr, int size)
>>>> +{
>>>> +    CXLType3Dev *ct3d = CT3(pci_dev);
>>>> +    PCIEDOE *doe = &ct3d->doe;
>>>> +    DOECap *doe_cap;
>>>> +
>>>> +    doe_cap = pcie_doe_covers_addr(doe, addr);
>>>> +    if (doe_cap) {
>>>> +        DOE_DBG(">> %s addr=%x, size=%x\n", __func__, addr, size);
>>>> +        return pcie_doe_read_config(doe_cap, addr, size);
>>>> +    } else {
>>>> +        return pci_default_read_config(pci_dev, addr, size);
>>>> +    }
>>>> +}
>>>> +
>>>> +static void ct3d_config_write(PCIDevice *pci_dev,
>>>> +                            uint32_t addr, uint32_t val, int size)
>>>> +{
>>>> +    CXLType3Dev *ct3d = CT3(pci_dev);
>>>> +    PCIEDOE *doe = &ct3d->doe;
>>>> +    DOECap *doe_cap;
>>>> +
>>>> +    doe_cap = pcie_doe_covers_addr(doe, addr);
>>>> +    if (doe_cap) {
>>>> +        DOE_DBG(">> %s addr=%x, val=%x, size=%x\n", __func__, addr, val, size);
>>>> +        pcie_doe_write_config(doe_cap, addr, val, size);
>>>> +    } else {
>>>> +        pci_default_write_config(pci_dev, addr, val, size);
>>>> +    }
>>>> +}
>>>> 
>>>> static void build_dvsecs(CXLType3Dev *ct3d)
>>>> {
>>>> @@ -210,6 +354,9 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>>>>    ComponentRegisters *regs = &cxl_cstate->crb;
>>>>    MemoryRegion *mr = &regs->component_registers;
>>>>    uint8_t *pci_conf = pci_dev->config;
>>>> +    unsigned short msix_num = 2;
>>>> +    int rc;
>>>> +    int i;
>>>> 
>>>>    if (!ct3d->cxl_dstate.pmem) {
>>>>        cxl_setup_memory(ct3d, errp);
>>>> @@ -239,6 +386,28 @@ static void ct3_realize(PCIDevice *pci_dev, Error **errp)
>>>>                     PCI_BASE_ADDRESS_SPACE_MEMORY |
>>>>                         PCI_BASE_ADDRESS_MEM_TYPE_64,
>>>>                     &ct3d->cxl_dstate.device_registers);
>>>> +
>>>> +    msix_init_exclusive_bar(pci_dev, msix_num, 4, NULL);
>>>> +    for (i = 0; i < msix_num; i++) {
>>>> +        msix_vector_use(pci_dev, i);
>>>> +    }
>>>> +
>>>> +    /* msi_init(pci_dev, 0x60, 16, true, false, NULL); */  
>>> 
>>> Tidy this up or parameterize it.
>>> 
>>>> +
>>>> +    pcie_doe_init(pci_dev, &ct3d->doe);
>>>> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x160, true, 0);  
>>> 
>>> check rc here.
>>> 
>>>> +    rc = pcie_cap_doe_add(&ct3d->doe, 0x190, true, 1);
>>>> +    if (rc) {
>>>> +        error_setg(errp, "fail to add DOE cap");
>>>> +        return;
>>>> +    }
>>>> +
>>>> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_COMPLIANCE,
>>>> +                               cxl_doe_compliance_rsp);
>>>> +    pcie_doe_register_protocol(&ct3d->doe, CXL_VENDOR_ID, CXL_DOE_TABLE_ACCESS,
>>>> +                               cxl_doe_cdat_rsp);
>>>> +
>>>> +    cxl_doe_cdat_init(cxl_cstate);
>>>> }
>>>> 
>>>> static uint64_t cxl_md_get_addr(const MemoryDeviceState *md)
>>>> @@ -357,6 +526,9 @@ static void ct3_class_init(ObjectClass *oc, void *data)
>>>>    DeviceClass *dc = DEVICE_CLASS(oc);
>>>>    PCIDeviceClass *pc = PCI_DEVICE_CLASS(oc);
>>>>    MemoryDeviceClass *mdc = MEMORY_DEVICE_CLASS(oc);
>>>> +
>>>> +    pc->config_write = ct3d_config_write;
>>>> +    pc->config_read = ct3d_config_read;
>>>>    CXLType3Class *cvc = CXL_TYPE3_DEV_CLASS(oc);
>>>> 
>>>>    pc->realize = ct3_realize;
>>>> diff --git a/include/hw/cxl/cxl_cdat.h b/include/hw/cxl/cxl_cdat.h
>>>> new file mode 100644
>>>> index 0000000..fbbd494
>>>> --- /dev/null
>>>> +++ b/include/hw/cxl/cxl_cdat.h
>>>> @@ -0,0 +1,120 @@
>>>> +#include "hw/cxl/cxl_pci.h"
>>>> +
>>>> +
>>>> +enum {
>>>> +    CXL_DOE_COMPLIANCE             = 0,
>>>> +    CXL_DOE_TABLE_ACCESS           = 2,
>>>> +    CXL_DOE_MAX_PROTOCOL
>>>> +};
>>>> +
>>>> +#define CXL_DOE_PROTOCOL_COMPLIANCE ((CXL_DOE_COMPLIANCE << 16) | CXL_VENDOR_ID)
>>>> +#define CXL_DOE_PROTOCOL_CDAT     ((CXL_DOE_TABLE_ACCESS << 16) | CXL_VENDOR_ID)
>>>> +
>>>> +/*
>>>> + * DOE CDAT Table Protocol (CXL Spec)
>>>> + */
>>>> +#define CXL_DOE_TAB_REQ 0
>>>> +#define CXL_DOE_TAB_RSP 0
>>>> +#define CXL_DOE_TAB_TYPE_CDAT 0
>>>> +#define CXL_DOE_TAB_ENT_MAX 0xFFFF
>>>> +
>>>> +/* Read Entry Request, 8.1.11.1 Table 134 */
>>>> +struct cxl_cdat {
>>>> +    DOEHeader header;
>>>> +    uint8_t req_code;
>>>> +    uint8_t table_type;
>>>> +    uint16_t entry_handle;
>>>> +} QEMU_PACKED;
>>>> +
>>>> +/* Read Entry Response, 8.1.11.1 Table 135 */
>>>> +#define cxl_cdat_rsp    cxl_cdat    /* Same as request */
>>>> +  
>>> ... Note I'm just snipping out these big defines as I check them
>>> against the spec :) Hence the gap.
>>> ...
>>> 
>>>> +struct cdat_dsmscis {
>>>> +    struct cdat_sub_header header;
>>>> +    uint8_t DSMASH_handle;  
>>> 
>>> DMSAS_handle;
>>> 
>>>> +    uint8_t reserved2[3];
>>>> +    uint64_t memory_side_cache_size;
>>>> +    uint32_t cache_attributes;
>>>> +} QEMU_PACKED;
>>>> +  
>>> 
>>> 
>>>> +
>>>> +struct cdat_sslbis_header {
>>>> +    struct cdat_sub_header header;
>>>> +    uint8_t data_type;
>>>> +    uint8_t reserved2[3];
>>>> +    uint64_t entry_base_unit;
>>>> +} QEMU_PACKED;
>>>> diff --git a/include/hw/cxl/cxl_compl.h b/include/hw/cxl/cxl_compl.h
>>>> new file mode 100644
>>>> index 0000000..ebbe488
>>>> --- /dev/null
>>>> +++ b/include/hw/cxl/cxl_compl.h
>>>> @@ -0,0 +1,289 @@
>>>> +/*
>>>> + * CXL Compliance Mode Protocol  
>>> 
>>> Needs license etc I think
>>> 
>>>> + */  
>>> 
>>> A bunch of the responses below are all of this form, perhaps we
>>> could just have one cxl_compliance_mode_generic_status_rsp ?
>>> if you really want to then define the others perhaps use
>>> #define to do it rather than lots of identical structures
>>> each specified fully.  
>> 
>> 
>> Got it.   Should have simplified it after cutting and pasting so many times.
>>> 
>>> 
>>>> +struct cxl_compliance_mode_inject_mac_delay_rsp {
>>>> +    DOEHeader header;
>>>> +    uint8_t rsp_code;
>>>> +    uint8_t version;
>>>> +    uint8_t length;
>>>> +    uint8_t status;
>>>> +} QEMU_PACKED;
>>>> +  
>>> 
>>> 
>>> 
>>>> +struct cxl_compliance_mode_ignore_bit_error {
>>>> +    DOEHeader header;
>>>> +    uint8_t req_code;
>>>> +    uint8_t version;
>>>> +    uint16_t reserved;
>>>> +    uint8_t opcode;
>>>> +} QEMU_PACKED;
>>>> +  
>>> This last one doesn't seem to line up with the CXL 2.0 spec.  
>> 
>> Good catch!
>> 
>>> 
>>>> +struct cxl_compliance_mode_ignore_bit_error_rsp {
>>>> +    DOEHeader header;
>>>> +    uint8_t rsp_code;
>>>> +    uint8_t version;
>>>> +    uint8_t reserved[6];
>>>> +} QEMU_PACKED;  
>>> 
>>> 
>>>> diff --git a/include/hw/cxl/cxl_component.h b/include/hw/cxl/cxl_component.h
>>>> index 762feb5..23923df 100644
>>>> --- a/include/hw/cxl/cxl_component.h
>>>> +++ b/include/hw/cxl/cxl_component.h
>>>> @@ -132,6 +132,23 @@ HDM_DECODER_INIT(0);
>>>> _Static_assert((CXL_SNOOP_REGISTERS_OFFSET + CXL_SNOOP_REGISTERS_SIZE) < 0x1000,
>>>>               "No space for registers");  
>>> 
>>> ...
>>> 
>>>> typedef struct component_registers {
>>>>    /*
>>>>     * Main memory region to be registered with QEMU core.
>>>> @@ -160,6 +177,10 @@ typedef struct component_registers {
>>>>    MemoryRegionOps *special_ops;
>>>> } ComponentRegisters;
>>>> 
>>>> +typedef struct cdat_struct {
>>>> +    void *base;
>>>> +    uint32_t length;
>>>> +} CDATStruct;
>>>> /*
>>>> * A CXL component represents all entities in a CXL hierarchy. This includes,
>>>> * host bridges, root ports, upstream/downstream switch ports, and devices
>>>> @@ -173,6 +194,104 @@ typedef struct cxl_component {
>>>>            struct PCIDevice *pdev;
>>>>        };
>>>>    };
>>>> +
>>>> +    /*
>>>> +     * SW write write mailbox and GO on last DW
>>>> +     * device sets READY of offset DW in DO types to copy
>>>> +     * to read mailbox register on subsequent cfg_read
>>>> +     * of read mailbox register and then on cfg_write to
>>>> +     * denote success read increments offset to next DW
>>>> +     */
>>>> +
>>>> +    union doe_request_u {
>>>> +        /* Compliance DOE Data Objects Type=0*/
>>>> +        struct cxl_compliance_mode_cap
>>>> +            mode_cap;  
>>> 
>>> I'd only add line breaks for the longer ones of these.
>>> 
>>>> +        struct cxl_compliance_mode_status
>>>> +            mode_status;
>>>> +        struct cxl_compliance_mode_halt
>>>> +            mode_halt;
>>>> +        struct cxl_compliance_mode_multiple_write_streaming
>>>> +            multiple_write_streaming;
>>>> +        struct cxl_compliance_mode_producer_consumer
>>>> +            producer_consumer;
>>>> +        struct cxl_compliance_mode_inject_bogus_writes
>>>> +            inject_bogus_writes;
>>>> +        struct cxl_compliance_mode_inject_poison
>>>> +            inject_poison;
>>>> +        struct cxl_compliance_mode_inject_crc
>>>> +            inject_crc;
>>>> +        struct cxl_compliance_mode_inject_flow_control
>>>> +            inject_flow_control;
>>>> +        struct cxl_compliance_mode_toggle_cache_flush
>>>> +            toggle_cache_flush;
>>>> +        struct cxl_compliance_mode_inject_mac_delay
>>>> +            inject_mac_delay;
>>>> +        struct cxl_compliance_mode_insert_unexp_mac
>>>> +            insert_unexp_mac;
>>>> +        struct cxl_compliance_mode_inject_viral
>>>> +            inject_viral;
>>>> +        struct cxl_compliance_mode_inject_almp
>>>> +            inject_almp;
>>>> +        struct cxl_compliance_mode_ignore_almp
>>>> +            ignore_almp;
>>>> +        struct cxl_compliance_mode_ignore_bit_error
>>>> +            ignore_bit_error;
>>>> +        char data_byte[128];
>>>> +    } doe_request;
>>>> +
>>>> +    union doe_resp_u {
>>>> +        /* Compliance DOE Data Objects Type=0*/
>>>> +        struct cxl_compliance_mode_cap_rsp
>>>> +            cap_rsp;
>>>> +        struct cxl_compliance_mode_status_rsp
>>>> +            status_rsp;
>>>> +        struct cxl_compliance_mode_halt_rsp
>>>> +            halt_rsp;
>>>> +        struct cxl_compliance_mode_multiple_write_streaming_rsp
>>>> +            multiple_write_streaming_rsp;
>>>> +        struct cxl_compliance_mode_producer_consumer_rsp
>>>> +            producer_consumer_rsp;
>>>> +        struct cxl_compliance_mode_inject_bogus_writes_rsp
>>>> +            inject_bogus_writes_rsp;
>>>> +        struct cxl_compliance_mode_inject_poison_rsp
>>>> +            inject_poison_rsp;
>>>> +        struct cxl_compliance_mode_inject_crc_rsp
>>>> +            inject_crc_rsp;
>>>> +        struct cxl_compliance_mode_inject_flow_control_rsp
>>>> +            inject_flow_control_rsp;
>>>> +        struct cxl_compliance_mode_toggle_cache_flush_rsp
>>>> +            toggle_cache_flush_rsp;
>>>> +        struct cxl_compliance_mode_inject_mac_delay_rsp
>>>> +            inject_mac_delay_rsp;
>>>> +        struct cxl_compliance_mode_insert_unexp_mac_rsp
>>>> +            insert_unexp_mac_rsp;
>>>> +        struct cxl_compliance_mode_inject_viral
>>>> +            inject_viral_rsp;
>>>> +        struct cxl_compliance_mode_inject_almp_rsp
>>>> +            inject_almp_rsp;
>>>> +        struct cxl_compliance_mode_ignore_almp_rsp
>>>> +            ignore_almp_rsp;
>>>> +        struct cxl_compliance_mode_ignore_bit_error_rsp
>>>> +            ignore_bit_error_rsp;
>>>> +        char data_byte[520 * 4];
>>>> +        uint32_t data_dword[520];
>>>> +    } doe_resp;
>>>> +
>>>> +    /* Table entry types */  
>>> Hmm. Not sure all CXL components will have CDAT.  Root ports for
>>> example?
>>> 
>>>> +    struct cdat_table_header cdat_header;
>>>> +    struct cdat_dsmas dsmas;
>>>> +    struct cdat_dslbis dslbis;  
>>> 
>>> As I said in v1, some of these will need to be multiples so this
>>> flat structure just doesn't work.
>>> 
>>>> +    struct cdat_dsmscis dsmscis;
>>>> +    struct cdat_dsis dsis;
>>>> +    struct cdat_dsemts dsemts;
>>>> +    struct {
>>>> +        struct cdat_sslbis_header sslbis_h;
>>>> +        struct cdat_sslbe sslbe[3];  
>>> 
>>> I'm curious.  Why 3?  This is between pairs of ports so 1USP 2DSP switch?
>>> 
>>>> +    } sslbis;  
>>> 
>>> 
>>>> +
>>>> +    CDATStruct *cdat_ent;
>>>> +    int cdat_ent_len;
>>>> } CXLComponentState;
>>>> 
>>>> void cxl_component_register_block_init(Object *obj,
>>>> @@ -184,4 +303,11 @@ void cxl_component_register_init_common(uint32_t *reg_state,
>>>> void cxl_component_create_dvsec(CXLComponentState *cxl_cstate, uint16_t length,
>>>>                                uint16_t type, uint8_t rev, uint8_t *body);
>>>> 
>>>> +void cxl_component_create_doe(CXLComponentState *cxl_cstate,
>>>> +                              uint16_t offset, unsigned vec);  
>>> 
>>> Doesn't seem to exist.
>>> 
>>> Some of the following are local to one file so shouldn't be here iether.
>>> 
>>>> +uint32_t cxl_doe_compliance_init(DOECap *doe_cap);
>>>> +bool cxl_doe_compliance_rsp(DOECap *doe_cap);
>>>> +void cxl_doe_cdat_init(CXLComponentState *cxl_cstate);
>>>> +bool cxl_doe_cdat_rsp(DOECap *doe_cap);
>>>> +uint32_t cdat_zero_checksum(uint32_t *mbox, uint32_t dw_cnt);  
>>> 
>>> Doesn't seem to exist.
>>> 
>>>> #endif
>>>> diff --git a/include/hw/cxl/cxl_device.h b/include/hw/cxl/cxl_device.h
>>>> index c608ced..08bf646 100644
>>>> --- a/include/hw/cxl/cxl_device.h
>>>> +++ b/include/hw/cxl/cxl_device.h
>>>> @@ -223,6 +223,9 @@ typedef struct cxl_type3_dev {
>>>>    /* State */
>>>>    CXLComponentState cxl_cstate;
>>>>    CXLDeviceState cxl_dstate;
>>>> +
>>>> +    /* DOE */
>>>> +    PCIEDOE doe;
>>>> } CXLType3Dev;
>>>> 
>>>> #ifndef TYPE_CXL_TYPE3_DEV
>>>> diff --git a/include/hw/cxl/cxl_pci.h b/include/hw/cxl/cxl_pci.h
>>>> index 9ec28c9..5cab197 100644
>>>> --- a/include/hw/cxl/cxl_pci.h
>>>> +++ b/include/hw/cxl/cxl_pci.h
>>>> @@ -12,6 +12,8 @@
>>>> 
>>>> #include "hw/pci/pci.h"
>>>> #include "hw/pci/pcie.h"
>>>> +#include "hw/cxl/cxl_cdat.h"
>>>> +#include "hw/cxl/cxl_compl.h"
>>>> 
>>>> #define CXL_VENDOR_ID 0x1e98
>>>> 
>>>> @@ -54,6 +56,8 @@ struct dvsec_header {
>>>> _Static_assert(sizeof(struct dvsec_header) == 10,
>>>>               "dvsec header size incorrect");
>>>> 
>>>> +/* CXL 2.0 - 8.1.11 */
>>>> +  
>>> 
>>> Clean this out next time - doesn't belong in this patch.
>>> 
>>>> /*
>>>> * CXL 2.0 devices must implement certain DVSEC IDs, and can [optionally]
>>>> * implement others.  
>>> 
>> 
>