powerpc/eeh: Output PHB3 diag-data
authorGavin Shan <shangw@linux.vnet.ibm.com>
Fri, 6 Sep 2013 01:00:04 +0000 (09:00 +0800)
committerBenjamin Herrenschmidt <benh@kernel.crashing.org>
Fri, 11 Oct 2013 05:54:12 +0000 (16:54 +1100)
The patch adds function ioda_eeh_phb3_phb_diag() to dump PHB3
PHB diag-data. That's called while detecting informative errors
or frozen PE on the specific PHB.

Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
arch/powerpc/include/asm/opal.h
arch/powerpc/platforms/powernv/eeh-ioda.c

index 51e3b26..4cc33ba 100644 (file)
@@ -460,10 +460,12 @@ enum {
 
 enum {
        OPAL_PHB_ERROR_DATA_TYPE_P7IOC = 1,
+       OPAL_PHB_ERROR_DATA_TYPE_PHB3 = 2
 };
 
 enum {
        OPAL_P7IOC_NUM_PEST_REGS = 128,
+       OPAL_PHB3_NUM_PEST_REGS = 256
 };
 
 struct OpalIoPhbErrorCommon {
@@ -531,6 +533,69 @@ struct OpalIoP7IOCPhbErrorData {
        uint64_t pestB[OPAL_P7IOC_NUM_PEST_REGS];
 };
 
+struct OpalIoPhb3ErrorData {
+       struct OpalIoPhbErrorCommon common;
+
+       uint32_t brdgCtl;
+
+       /* PHB3 UTL regs */
+       uint32_t portStatusReg;
+       uint32_t rootCmplxStatus;
+       uint32_t busAgentStatus;
+
+       /* PHB3 cfg regs */
+       uint32_t deviceStatus;
+       uint32_t slotStatus;
+       uint32_t linkStatus;
+       uint32_t devCmdStatus;
+       uint32_t devSecStatus;
+
+       /* cfg AER regs */
+       uint32_t rootErrorStatus;
+       uint32_t uncorrErrorStatus;
+       uint32_t corrErrorStatus;
+       uint32_t tlpHdr1;
+       uint32_t tlpHdr2;
+       uint32_t tlpHdr3;
+       uint32_t tlpHdr4;
+       uint32_t sourceId;
+
+       uint32_t rsv3;
+
+       /* Record data about the call to allocate a buffer */
+       uint64_t errorClass;
+       uint64_t correlator;
+
+       uint64_t nFir;                  /* 000 */
+       uint64_t nFirMask;              /* 003 */
+       uint64_t nFirWOF;               /* 008 */
+
+       /* PHB3 MMIO Error Regs */
+       uint64_t phbPlssr;              /* 120 */
+       uint64_t phbCsr;                /* 110 */
+       uint64_t lemFir;                /* C00 */
+       uint64_t lemErrorMask;          /* C18 */
+       uint64_t lemWOF;                /* C40 */
+       uint64_t phbErrorStatus;        /* C80 */
+       uint64_t phbFirstErrorStatus;   /* C88 */
+       uint64_t phbErrorLog0;          /* CC0 */
+       uint64_t phbErrorLog1;          /* CC8 */
+       uint64_t mmioErrorStatus;       /* D00 */
+       uint64_t mmioFirstErrorStatus;  /* D08 */
+       uint64_t mmioErrorLog0;         /* D40 */
+       uint64_t mmioErrorLog1;         /* D48 */
+       uint64_t dma0ErrorStatus;       /* D80 */
+       uint64_t dma0FirstErrorStatus;  /* D88 */
+       uint64_t dma0ErrorLog0;         /* DC0 */
+       uint64_t dma0ErrorLog1;         /* DC8 */
+       uint64_t dma1ErrorStatus;       /* E00 */
+       uint64_t dma1FirstErrorStatus;  /* E08 */
+       uint64_t dma1ErrorLog0;         /* E40 */
+       uint64_t dma1ErrorLog1;         /* E48 */
+       uint64_t pestA[OPAL_PHB3_NUM_PEST_REGS];
+       uint64_t pestB[OPAL_PHB3_NUM_PEST_REGS];
+};
+
 typedef struct oppanel_line {
        const char *    line;
        uint64_t        line_len;
index f426f4c..02245ce 100644 (file)
@@ -747,6 +747,73 @@ static void ioda_eeh_p7ioc_phb_diag(struct pci_controller *hose,
        }
 }
 
+static void ioda_eeh_phb3_phb_diag(struct pci_controller *hose,
+                                   struct OpalIoPhbErrorCommon *common)
+{
+       struct OpalIoPhb3ErrorData *data;
+       int i;
+
+       data = (struct OpalIoPhb3ErrorData*)common;
+       pr_info("PHB3 PHB#%x Diag-data (Version: %d)\n\n",
+               hose->global_number, common->version);
+
+       pr_info("  brdgCtl:              %08x\n", data->brdgCtl);
+
+       pr_info("  portStatusReg:        %08x\n", data->portStatusReg);
+       pr_info("  rootCmplxStatus:      %08x\n", data->rootCmplxStatus);
+       pr_info("  busAgentStatus:       %08x\n", data->busAgentStatus);
+
+       pr_info("  deviceStatus:         %08x\n", data->deviceStatus);
+       pr_info("  slotStatus:           %08x\n", data->slotStatus);
+       pr_info("  linkStatus:           %08x\n", data->linkStatus);
+       pr_info("  devCmdStatus:         %08x\n", data->devCmdStatus);
+       pr_info("  devSecStatus:         %08x\n", data->devSecStatus);
+
+       pr_info("  rootErrorStatus:      %08x\n", data->rootErrorStatus);
+       pr_info("  uncorrErrorStatus:    %08x\n", data->uncorrErrorStatus);
+       pr_info("  corrErrorStatus:      %08x\n", data->corrErrorStatus);
+       pr_info("  tlpHdr1:              %08x\n", data->tlpHdr1);
+       pr_info("  tlpHdr2:              %08x\n", data->tlpHdr2);
+       pr_info("  tlpHdr3:              %08x\n", data->tlpHdr3);
+       pr_info("  tlpHdr4:              %08x\n", data->tlpHdr4);
+       pr_info("  sourceId:             %08x\n", data->sourceId);
+       pr_info("  errorClass:           %016llx\n", data->errorClass);
+       pr_info("  correlator:           %016llx\n", data->correlator);
+       pr_info("  nFir:                 %016llx\n", data->nFir);
+       pr_info("  nFirMask:             %016llx\n", data->nFirMask);
+       pr_info("  nFirWOF:              %016llx\n", data->nFirWOF);
+       pr_info("  PhbPlssr:             %016llx\n", data->phbPlssr);
+       pr_info("  PhbCsr:               %016llx\n", data->phbCsr);
+       pr_info("  lemFir:               %016llx\n", data->lemFir);
+       pr_info("  lemErrorMask:         %016llx\n", data->lemErrorMask);
+       pr_info("  lemWOF:               %016llx\n", data->lemWOF);
+       pr_info("  phbErrorStatus:       %016llx\n", data->phbErrorStatus);
+       pr_info("  phbFirstErrorStatus:  %016llx\n", data->phbFirstErrorStatus);
+       pr_info("  phbErrorLog0:         %016llx\n", data->phbErrorLog0);
+       pr_info("  phbErrorLog1:         %016llx\n", data->phbErrorLog1);
+       pr_info("  mmioErrorStatus:      %016llx\n", data->mmioErrorStatus);
+       pr_info("  mmioFirstErrorStatus: %016llx\n", data->mmioFirstErrorStatus);
+       pr_info("  mmioErrorLog0:        %016llx\n", data->mmioErrorLog0);
+       pr_info("  mmioErrorLog1:        %016llx\n", data->mmioErrorLog1);
+       pr_info("  dma0ErrorStatus:      %016llx\n", data->dma0ErrorStatus);
+       pr_info("  dma0FirstErrorStatus: %016llx\n", data->dma0FirstErrorStatus);
+       pr_info("  dma0ErrorLog0:        %016llx\n", data->dma0ErrorLog0);
+       pr_info("  dma0ErrorLog1:        %016llx\n", data->dma0ErrorLog1);
+       pr_info("  dma1ErrorStatus:      %016llx\n", data->dma1ErrorStatus);
+       pr_info("  dma1FirstErrorStatus: %016llx\n", data->dma1FirstErrorStatus);
+       pr_info("  dma1ErrorLog0:        %016llx\n", data->dma1ErrorLog0);
+       pr_info("  dma1ErrorLog1:        %016llx\n", data->dma1ErrorLog1);
+
+       for (i = 0; i < OPAL_PHB3_NUM_PEST_REGS; i++) {
+               if ((data->pestA[i] >> 63) == 0 &&
+                   (data->pestB[i] >> 63) == 0)
+                       continue;
+
+               pr_info("  PE[%3d] PESTA:        %016llx\n", i, data->pestA[i]);
+               pr_info("          PESTB:        %016llx\n", data->pestB[i]);
+       }
+}
+
 static void ioda_eeh_phb_diag(struct pci_controller *hose)
 {
        struct pnv_phb *phb = hose->private_data;
@@ -765,6 +832,9 @@ static void ioda_eeh_phb_diag(struct pci_controller *hose)
        case OPAL_PHB_ERROR_DATA_TYPE_P7IOC:
                ioda_eeh_p7ioc_phb_diag(hose, common);
                break;
+       case OPAL_PHB_ERROR_DATA_TYPE_PHB3:
+               ioda_eeh_phb3_phb_diag(hose, common);
+               break;
        default:
                pr_warning("%s: Unrecognized I/O chip %d\n",
                           __func__, common->ioType);