Browse Source

PCI/AER: Add TLP header information to tracepoint

When a PCIe AER error occurs, the TLP header information is printed in the
kernel message but it is missing from the tracepoint.  A userspace program
can use this information in the tracepoint to better analyze problems.

To enable the tracepoint:

  echo 1 > /sys/kernel/debug/tracing/events/ras/aer_event/enable

Example tracepoint output:

  $ cat /sys/kernel/debug/tracing/trace
  aer_event: 0000:01:00.0
  PCIe Bus Error: severity=Uncorrected, non-fatal, Completer Abort
  TLP Header={0x0,0x1,0x2,0x3}

Signed-off-by: Thomas Tai <thomas.tai@oracle.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Thomas Tai 7 years ago
parent
commit
2af8641b2a
2 changed files with 20 additions and 6 deletions
  1. 2 2
      drivers/pci/pcie/aer/aerdrv_errprint.c
  2. 18 4
      include/ras/ras_event.h

+ 2 - 2
drivers/pci/pcie/aer/aerdrv_errprint.c

@@ -189,7 +189,7 @@ out:
 		pci_err(dev, "  Error of this Agent(%04x) is reported first\n", id);
 
 	trace_aer_event(dev_name(&dev->dev), (info->status & ~info->mask),
-			info->severity);
+			info->severity, info->tlp_header_valid, &info->tlp);
 }
 
 void aer_print_port_info(struct pci_dev *dev, struct aer_err_info *info)
@@ -251,6 +251,6 @@ void cper_print_aer(struct pci_dev *dev, int aer_severity,
 		__print_tlp_header(dev, &aer->header_log);
 
 	trace_aer_event(dev_name(&dev->dev), (status & ~mask),
-			aer_severity);
+			aer_severity, tlp_header_valid, &aer->header_log);
 }
 #endif

+ 18 - 4
include/ras/ras_event.h

@@ -298,30 +298,44 @@ TRACE_EVENT(non_standard_event,
 TRACE_EVENT(aer_event,
 	TP_PROTO(const char *dev_name,
 		 const u32 status,
-		 const u8 severity),
+		 const u8 severity,
+		 const u8 tlp_header_valid,
+		 struct aer_header_log_regs *tlp),
 
-	TP_ARGS(dev_name, status, severity),
+	TP_ARGS(dev_name, status, severity, tlp_header_valid, tlp),
 
 	TP_STRUCT__entry(
 		__string(	dev_name,	dev_name	)
 		__field(	u32,		status		)
 		__field(	u8,		severity	)
+		__field(	u8, 		tlp_header_valid)
+		__array(	u32, 		tlp_header, 4	)
 	),
 
 	TP_fast_assign(
 		__assign_str(dev_name, dev_name);
 		__entry->status		= status;
 		__entry->severity	= severity;
+		__entry->tlp_header_valid = tlp_header_valid;
+		if (tlp_header_valid) {
+			__entry->tlp_header[0] = tlp->dw0;
+			__entry->tlp_header[1] = tlp->dw1;
+			__entry->tlp_header[2] = tlp->dw2;
+			__entry->tlp_header[3] = tlp->dw3;
+		}
 	),
 
-	TP_printk("%s PCIe Bus Error: severity=%s, %s\n",
+	TP_printk("%s PCIe Bus Error: severity=%s, %s, TLP Header=%s\n",
 		__get_str(dev_name),
 		__entry->severity == AER_CORRECTABLE ? "Corrected" :
 			__entry->severity == AER_FATAL ?
 			"Fatal" : "Uncorrected, non-fatal",
 		__entry->severity == AER_CORRECTABLE ?
 		__print_flags(__entry->status, "|", aer_correctable_errors) :
-		__print_flags(__entry->status, "|", aer_uncorrectable_errors))
+		__print_flags(__entry->status, "|", aer_uncorrectable_errors),
+		__entry->tlp_header_valid ?
+			__print_array(__entry->tlp_header, 4, 4) :
+			"Not available")
 );
 
 /*