Module openj9.dtfj

Package com.ibm.j9ddr.tools.ddrinteractive.commands

Class TraceFileHeaderWriter

java.lang.Object

com.ibm.j9ddr.tools.ddrinteractive.commands.TraceFileHeaderWriter

public class TraceFileHeaderWriter
extends Object

Constructor Summary

Constructors

Constructor	Description
TraceFileHeaderWriter(String snapTraceFileName, boolean bigEndian, int cpuCount, int wordSize, int bufferSize, String arch, String processorSubType, String serviceLevel, String startupOptions, String traceConfig, long startPlatform, long startSystem, int type, int generations)	This class generates a trace file header.

Method Summary

Modifier and Type	Method	Description
byte[]	createAndWriteTraceFileHeader()	Reassembles the trace file header from scratch (because there wasn't one in the core dump we could reuse).

Methods declared in class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

TraceFileHeaderWriter

public TraceFileHeaderWriter(String snapTraceFileName,
                             boolean bigEndian,
                             int cpuCount,
                             int wordSize,
                             int bufferSize,
                             String arch,
                             String processorSubType,
                             String serviceLevel,
                             String startupOptions,
                             String traceConfig,
                             long startPlatform,
                             long startSystem,
                             int type,
                             int generations)
                      throws IOException

This class generates a trace file header. It is intended to be used when a core dump has been taken before the JVM has written out any trace, the trace header is not initialized until that happens. There are a lot of arguments to pass but this data can be obtained from the dump and prevents this code needing to exist in 23, 24 and 26 flavours.

Parameters:

snapTraceFileName - the filename to write the trace header to

bigEndian - true if the dump came from a big endian system.

cpuCount - the number of cpu's on the system the dump came from

wordSize - the wordsize of the system the dump came from

bufferSize - the size of the trace buffers in the dump

arch - a string denoting the architecture of the system the dump came from

processorSubType - a string denoting the cpu subtype of the system the dump came from

serviceLevel - the service level of the jvm the dump came from

startupOptions - a string containing the jvm startup options

generations -

type -

Throws:

IOException

Method Detail

createAndWriteTraceFileHeader

public byte[] createAndWriteTraceFileHeader()
                                     throws IOException

Reassembles the trace file header from scratch (because there wasn't one in the core dump we could reuse). Based on initTraceHeader in ut_trace.c

Throws:

IOException