powerpc/eeh: Do probe on pci_dn

Originally, EEH core probes on device_node or pci_dev to populate
EEH devices and PEs, which conflicts with the fact: SRIOV VFs are
usually enabled and created by PF's driver and they don't have the
corresponding device_nodes. Instead, SRIOV VFs have dynamically
created pci_dn, which can be used for EEH probe.

The patch reworks EEH probe for PowerNV and pSeries platforms to
do probing based on pci_dn, instead of pci_dev or device_node any
more.

Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

3 files changed, 143 insertions, 87 deletions

diff --git a/arch/powerpc/platforms/powernv/eeh-powernv.c b/arch/powerpc/platforms/powernv/eeh-powernv.c
index 8eac8c57ee86..dcc524fe2a30 100644
--- a/arch/powerpc/platforms/powernv/eeh-powernv.c
+++ b/arch/powerpc/platforms/powernv/eeh-powernv.c
@@ -286,10 +286,82 @@ static int pnv_eeh_post_init(void)
         return ret;
 }
 
+static int pnv_eeh_cap_start(struct pci_dn *pdn)
+{
+        u32 status;
+
+        if (!pdn)
+                return 0;
+
+        pnv_pci_cfg_read(pdn, PCI_STATUS, 2, &status);
+        if (!(status & PCI_STATUS_CAP_LIST))
+                return 0;
+
+        return PCI_CAPABILITY_LIST;
+}
+
+static int pnv_eeh_find_cap(struct pci_dn *pdn, int cap)
+{
+        int pos = pnv_eeh_cap_start(pdn);
+        int cnt = 48;   /* Maximal number of capabilities */
+        u32 id;
+
+        if (!pos)
+                return 0;
+
+        while (cnt--) {
+                pnv_pci_cfg_read(pdn, pos, 1, &pos);
+                if (pos < 0x40)
+                        break;
+
+                pos &= ~3;
+                pnv_pci_cfg_read(pdn, pos + PCI_CAP_LIST_ID, 1, &id);
+                if (id == 0xff)
+                        break;
+
+                /* Found */
+                if (id == cap)
+                        return pos;
+
+                /* Next one */
+                pos += PCI_CAP_LIST_NEXT;
+        }
+
+        return 0;
+}
+
+static int pnv_eeh_find_ecap(struct pci_dn *pdn, int cap)
+{
+        struct eeh_dev *edev = pdn_to_eeh_dev(pdn);
+        u32 header;
+        int pos = 256, ttl = (4096 - 256) / 8;
+
+        if (!edev || !edev->pcie_cap)
+                return 0;
+        if (pnv_pci_cfg_read(pdn, pos, 4, &header) != PCIBIOS_SUCCESSFUL)
+                return 0;
+        else if (!header)
+                return 0;
+
+        while (ttl-- > 0) {
+                if (PCI_EXT_CAP_ID(header) == cap && pos)
+                        return pos;
+
+                pos = PCI_EXT_CAP_NEXT(header);
+                if (pos < 256)
+                        break;
+
+                if (pnv_pci_cfg_read(pdn, pos, 4, &header) != PCIBIOS_SUCCESSFUL)
+                        break;
+        }
+
+        return 0;
+}
+
 /**
- * pnv_eeh_dev_probe - Do probe on PCI device
- * @dev: PCI device
- * @flag: unused
+ * pnv_eeh_probe - Do probe on PCI device
+ * @pdn: PCI device node
+ * @data: unused
  *
  * When EEH module is installed during system boot, all PCI devices
  * are checked one by one to see if it supports EEH. The function
@@ -303,12 +375,12 @@ static int pnv_eeh_post_init(void)
  * was possiblly triggered by EEH core, the binding between EEH device
  * and the PCI device isn't built yet.
  */
-static int pnv_eeh_dev_probe(struct pci_dev *dev, void *flag)
+static void *pnv_eeh_probe(struct pci_dn *pdn, void *data)
 {
-        struct pci_controller *hose = pci_bus_to_host(dev->bus);
+        struct pci_controller *hose = pdn->phb;
         struct pnv_phb *phb = hose->private_data;
-        struct device_node *dn = pci_device_to_OF_node(dev);
-        struct eeh_dev *edev = of_node_to_eeh_dev(dn);
+        struct eeh_dev *edev = pdn_to_eeh_dev(pdn);
+        uint32_t pcie_flags;
         int ret;
 
         /*
@@ -317,40 +389,42 @@ static int pnv_eeh_dev_probe(struct pci_dev *dev, void *flag)
          * the root bridge. So it's not reasonable to continue
          * the probing.
          */
-        if (!dn || !edev || edev->pe)
-                return 0;
+        if (!edev || edev->pe)
+                return NULL;
 
         /* Skip for PCI-ISA bridge */
-        if ((dev->class >> 8) == PCI_CLASS_BRIDGE_ISA)
-                return 0;
+        if ((pdn->class_code >> 8) == PCI_CLASS_BRIDGE_ISA)
+                return NULL;
 
         /* Initialize eeh device */
-        edev->class_code = dev->class;
+        edev->class_code = pdn->class_code;
         edev->mode      &= 0xFFFFFF00;
-        if (dev->hdr_type == PCI_HEADER_TYPE_BRIDGE)
+        edev->pcix_cap = pnv_eeh_find_cap(pdn, PCI_CAP_ID_PCIX);
+        edev->pcie_cap = pnv_eeh_find_cap(pdn, PCI_CAP_ID_EXP);
+        edev->aer_cap  = pnv_eeh_find_ecap(pdn, PCI_EXT_CAP_ID_ERR);
+        if ((edev->class_code >> 8) == PCI_CLASS_BRIDGE_PCI) {
                 edev->mode |= EEH_DEV_BRIDGE;
-        edev->pcix_cap = pci_find_capability(dev, PCI_CAP_ID_PCIX);
-        if (pci_is_pcie(dev)) {
-                edev->pcie_cap = pci_pcie_cap(dev);
-
-                if (pci_pcie_type(dev) == PCI_EXP_TYPE_ROOT_PORT)
-                        edev->mode |= EEH_DEV_ROOT_PORT;
-                else if (pci_pcie_type(dev) == PCI_EXP_TYPE_DOWNSTREAM)
-                        edev->mode |= EEH_DEV_DS_PORT;
-
-                edev->aer_cap = pci_find_ext_capability(dev,
-                                                        PCI_EXT_CAP_ID_ERR);
+                if (edev->pcie_cap) {
+                        pnv_pci_cfg_read(pdn, edev->pcie_cap + PCI_EXP_FLAGS,
+                                         2, &pcie_flags);
+                        pcie_flags = (pcie_flags & PCI_EXP_FLAGS_TYPE) >> 4;
+                        if (pcie_flags == PCI_EXP_TYPE_ROOT_PORT)
+                                edev->mode |= EEH_DEV_ROOT_PORT;
+                        else if (pcie_flags == PCI_EXP_TYPE_DOWNSTREAM)
+                                edev->mode |= EEH_DEV_DS_PORT;
+                }
         }
 
-        edev->config_addr       = ((dev->bus->number << 8) | dev->devfn);
-        edev->pe_config_addr    = phb->bdfn_to_pe(phb, dev->bus, dev->devfn & 0xff);
+        edev->config_addr    = (pdn->busno << 8) | (pdn->devfn);
+        edev->pe_config_addr = phb->ioda.pe_rmap[edev->config_addr];
 
         /* Create PE */
         ret = eeh_add_to_parent_pe(edev);
         if (ret) {
-                pr_warn("%s: Can't add PCI dev %s to parent PE (%d)\n",
-                        __func__, pci_name(dev), ret);
-                return ret;
+                pr_warn("%s: Can't add PCI dev %04x:%02x:%02x.%01x to parent PE (%d)\n",
+                        __func__, hose->global_number, pdn->busno,
+                        PCI_SLOT(pdn->devfn), PCI_FUNC(pdn->devfn), ret);
+                return NULL;
         }
 
         /*
@@ -369,8 +443,10 @@ static int pnv_eeh_dev_probe(struct pci_dev *dev, void *flag)
          * Broadcom Austin 4-ports NICs (14e4:1657)
          * Broadcom Shiner 2-ports 10G NICs (14e4:168e)
          */
-        if ((dev->vendor == PCI_VENDOR_ID_BROADCOM && dev->device == 0x1657) ||
-            (dev->vendor == PCI_VENDOR_ID_BROADCOM && dev->device == 0x168e))
+        if ((pdn->vendor_id == PCI_VENDOR_ID_BROADCOM &&
+             pdn->device_id == 0x1657) ||
+            (pdn->vendor_id == PCI_VENDOR_ID_BROADCOM &&
+             pdn->device_id == 0x168e))
                 edev->pe->state |= EEH_PE_CFG_RESTRICTED;
 
         /*
@@ -380,7 +456,8 @@ static int pnv_eeh_dev_probe(struct pci_dev *dev, void *flag)
          * to PE reset.
          */
         if (!edev->pe->bus)
-                edev->pe->bus = dev->bus;
+                edev->pe->bus = pci_find_bus(hose->global_number,
+                                             pdn->busno);
 
         /*
          * Enable EEH explicitly so that we will do EEH check
@@ -391,7 +468,7 @@ static int pnv_eeh_dev_probe(struct pci_dev *dev, void *flag)
         /* Save memory bars */
         eeh_save_bars(edev);
 
-        return 0;
+        return NULL;
 }
 
 /**
@@ -1432,8 +1509,7 @@ static struct eeh_ops pnv_eeh_ops = {
         .name                   = "powernv",
         .init                   = pnv_eeh_init,
         .post_init              = pnv_eeh_post_init,
-        .of_probe               = NULL,
-        .dev_probe              = pnv_eeh_dev_probe,
+        .probe                  = pnv_eeh_probe,
         .set_option             = pnv_eeh_set_option,
         .get_pe_addr            = pnv_eeh_get_pe_addr,
         .get_state              = pnv_eeh_get_state,
diff --git a/arch/powerpc/platforms/pseries/eeh_pseries.c b/arch/powerpc/platforms/pseries/eeh_pseries.c
index a6c7e19f5eb3..a2946f72d5e7 100644
--- a/arch/powerpc/platforms/pseries/eeh_pseries.c
+++ b/arch/powerpc/platforms/pseries/eeh_pseries.c
@@ -118,9 +118,8 @@ static int pseries_eeh_init(void)
         return 0;
 }
 
-static int pseries_eeh_cap_start(struct device_node *dn)
+static int pseries_eeh_cap_start(struct pci_dn *pdn)
 {
-        struct pci_dn *pdn = PCI_DN(dn);
         u32 status;
 
         if (!pdn)
@@ -134,10 +133,9 @@ static int pseries_eeh_cap_start(struct device_node *dn)
 }
 
 
-static int pseries_eeh_find_cap(struct device_node *dn, int cap)
+static int pseries_eeh_find_cap(struct pci_dn *pdn, int cap)
 {
-        struct pci_dn *pdn = PCI_DN(dn);
-        int pos = pseries_eeh_cap_start(dn);
+        int pos = pseries_eeh_cap_start(pdn);
         int cnt = 48;   /* Maximal number of capabilities */
         u32 id;
 
@@ -160,10 +158,9 @@ static int pseries_eeh_find_cap(struct device_node *dn, int cap)
         return 0;
 }
 
-static int pseries_eeh_find_ecap(struct device_node *dn, int cap)
+static int pseries_eeh_find_ecap(struct pci_dn *pdn, int cap)
 {
-        struct pci_dn *pdn = PCI_DN(dn);
-        struct eeh_dev *edev = of_node_to_eeh_dev(dn);
+        struct eeh_dev *edev = pdn_to_eeh_dev(pdn);
         u32 header;
         int pos = 256;
         int ttl = (4096 - 256) / 8;
@@ -191,53 +188,44 @@ static int pseries_eeh_find_ecap(struct device_node *dn, int cap)
 }
 
 /**
- * pseries_eeh_of_probe - EEH probe on the given device
- * @dn: OF node
- * @flag: Unused
+ * pseries_eeh_probe - EEH probe on the given device
+ * @pdn: PCI device node
+ * @data: Unused
  *
  * When EEH module is installed during system boot, all PCI devices
  * are checked one by one to see if it supports EEH. The function
  * is introduced for the purpose.
  */
-static void *pseries_eeh_of_probe(struct device_node *dn, void *flag)
+static void *pseries_eeh_probe(struct pci_dn *pdn, void *data)
 {
         struct eeh_dev *edev;
         struct eeh_pe pe;
-        struct pci_dn *pdn = PCI_DN(dn);
-        const __be32 *classp, *vendorp, *devicep;
-        u32 class_code;
-        const __be32 *regs;
         u32 pcie_flags;
         int enable = 0;
         int ret;
 
         /* Retrieve OF node and eeh device */
-        edev = of_node_to_eeh_dev(dn);
-        if (edev->pe || !of_device_is_available(dn))
+        edev = pdn_to_eeh_dev(pdn);
+        if (!edev || edev->pe)
                 return NULL;
 
-        /* Retrieve class/vendor/device IDs */
-        classp = of_get_property(dn, "class-code", NULL);
-        vendorp = of_get_property(dn, "vendor-id", NULL);
-        devicep = of_get_property(dn, "device-id", NULL);
-
-        /* Skip for bad OF node or PCI-ISA bridge */
-        if (!classp || !vendorp || !devicep)
-                return NULL;
-        if (dn->type && !strcmp(dn->type, "isa"))
+        /* Check class/vendor/device IDs */
+        if (!pdn->vendor_id || !pdn->device_id || !pdn->class_code)
                 return NULL;
 
-        class_code = of_read_number(classp, 1);
+        /* Skip for PCI-ISA bridge */
+        if ((pdn->class_code >> 8) == PCI_CLASS_BRIDGE_ISA)
+                return NULL;
 
         /*
          * Update class code and mode of eeh device. We need
          * correctly reflects that current device is root port
          * or PCIe switch downstream port.
          */
-        edev->class_code = class_code;
-        edev->pcix_cap = pseries_eeh_find_cap(dn, PCI_CAP_ID_PCIX);
-        edev->pcie_cap = pseries_eeh_find_cap(dn, PCI_CAP_ID_EXP);
-        edev->aer_cap = pseries_eeh_find_ecap(dn, PCI_EXT_CAP_ID_ERR);
+        edev->class_code = pdn->class_code;
+        edev->pcix_cap = pseries_eeh_find_cap(pdn, PCI_CAP_ID_PCIX);
+        edev->pcie_cap = pseries_eeh_find_cap(pdn, PCI_CAP_ID_EXP);
+        edev->aer_cap = pseries_eeh_find_ecap(pdn, PCI_EXT_CAP_ID_ERR);
         edev->mode &= 0xFFFFFF00;
         if ((edev->class_code >> 8) == PCI_CLASS_BRIDGE_PCI) {
                 edev->mode |= EEH_DEV_BRIDGE;
@@ -252,24 +240,16 @@ static void *pseries_eeh_of_probe(struct device_node *dn, void *flag)
                 }
         }
 
-        /* Retrieve the device address */
-        regs = of_get_property(dn, "reg", NULL);
-        if (!regs) {
-                pr_warn("%s: OF node property %s::reg not found\n",
-                        __func__, dn->full_name);
-                return NULL;
-        }
-
         /* Initialize the fake PE */
         memset(&pe, 0, sizeof(struct eeh_pe));
         pe.phb = edev->phb;
-        pe.config_addr = of_read_number(regs, 1);
+        pe.config_addr = (pdn->busno << 16) | (pdn->devfn << 8);
 
         /* Enable EEH on the device */
         ret = eeh_ops->set_option(&pe, EEH_OPT_ENABLE);
         if (!ret) {
-                edev->config_addr = of_read_number(regs, 1);
                 /* Retrieve PE address */
+                edev->config_addr = (pdn->busno << 16) | (pdn->devfn << 8);
                 edev->pe_config_addr = eeh_ops->get_pe_addr(&pe);
                 pe.addr = edev->pe_config_addr;
 
@@ -285,16 +265,17 @@ static void *pseries_eeh_of_probe(struct device_node *dn, void *flag)
                         eeh_add_flag(EEH_ENABLED);
                         eeh_add_to_parent_pe(edev);
 
-                        pr_debug("%s: EEH enabled on %s PHB#%d-PE#%x, config addr#%x\n",
-                                __func__, dn->full_name, pe.phb->global_number,
-                                pe.addr, pe.config_addr);
-                } else if (dn->parent && of_node_to_eeh_dev(dn->parent) &&
-                           (of_node_to_eeh_dev(dn->parent))->pe) {
+                        pr_debug("%s: EEH enabled on %02x:%02x.%01x PHB#%d-PE#%x\n",
+                                __func__, pdn->busno, PCI_SLOT(pdn->devfn),
+                                PCI_FUNC(pdn->devfn), pe.phb->global_number,
+                                pe.addr);
+                } else if (pdn->parent && pdn_to_eeh_dev(pdn->parent) &&
+                           (pdn_to_eeh_dev(pdn->parent))->pe) {
                         /* This device doesn't support EEH, but it may have an
                          * EEH parent, in which case we mark it as supported.
                          */
-                        edev->config_addr = of_node_to_eeh_dev(dn->parent)->config_addr;
-                        edev->pe_config_addr = of_node_to_eeh_dev(dn->parent)->pe_config_addr;
+                        edev->config_addr = pdn_to_eeh_dev(pdn->parent)->config_addr;
+                        edev->pe_config_addr = pdn_to_eeh_dev(pdn->parent)->pe_config_addr;
                         eeh_add_to_parent_pe(edev);
                 }
         }
@@ -707,8 +688,7 @@ static int pseries_eeh_write_config(struct device_node *dn, int where, int size,
 static struct eeh_ops pseries_eeh_ops = {
         .name                   = "pseries",
         .init                   = pseries_eeh_init,
-        .of_probe               = pseries_eeh_of_probe,
-        .dev_probe              = NULL,
+        .probe                  = pseries_eeh_probe,
         .set_option             = pseries_eeh_set_option,
         .get_pe_addr            = pseries_eeh_get_pe_addr,
         .get_state              = pseries_eeh_get_state,
diff --git a/arch/powerpc/platforms/pseries/pci_dlpar.c b/arch/powerpc/platforms/pseries/pci_dlpar.c
index 89e23811199c..f735f4fee48c 100644
--- a/arch/powerpc/platforms/pseries/pci_dlpar.c
+++ b/arch/powerpc/platforms/pseries/pci_dlpar.c
@@ -82,7 +82,7 @@ struct pci_controller *init_phb_dynamic(struct device_node *dn)
         eeh_dev_phb_init_dynamic(phb);
 
         if (dn->child)
-                eeh_add_device_tree_early(dn);
+                eeh_add_device_tree_early(PCI_DN(dn));
 
         pcibios_scan_phb(phb);
         pcibios_finish_adding_to_bus(phb->bus);