Course Description

This course provides the foundations to perform crash dump analysis on HP OpenVMS Integrity servers. Itanium architectural characteristics are introduced in order to analyze crash dumps and system hangs. The calling standard is reviewed, allowing students to walk call frames to determine where improper arguments to a function call have been passed. Interrupts and exceptions are discussed, as well as their handling by the Software Interrupt Services (SWIS). Considerable time is spent exploring different types of crashes and reviewing specific examples.

Who Should Attend

hp OpenVMS System Programmers and Advanced Troubleshooters, especially those who understand the coding considerations for code that will run in kernel or executive mode.

Prerequisites

To get the most out of this course, students should have:

• Completion of OpenVMS Academy Programming Features - course code VM310
• The ability to read and ideally write code in VAX Macro
• The ability to program in the DEC C and Bliss languages
• Completion of OpenVMS Alpha Internals I - course code U3719S
• Completion of OpenVMS Alpha Internals II - course code U3720S
• Completion of OpenVMS Alpha Internals III - course code U3721S

Benefits of Attending this Class

Students attending this course should leave with the ability to:

• Identify the general Integrity architecture and procedure call types
• Describe what causes a crash to occur
• Describe the general characteristics of interruptions on Integrity servers
• Identify the reasons for an "Unexpected System Service Exception" crash
• Describe and analyze basic "Invalid Exception" crashes
• Identify and analyze basic "Page Fault with IPL too high" crashes
• Describe monitoring oriented system parameters
• Describe and perform basic analysis of the "kernel stack not valid" crash
• Identify general causes of SMP-related crashes
• Describe and analyze basic AST-related crashes
• Analyze system hangs

Course Contents

Overview of General Itanium Architecture

• General IA64 architecture
• Itanium application registers
• Predication
• Register conventions and usage considerations
• Using map and listing files

Procedure Calls

• OpenVMS calling standard on IA64
• Calling procedures
• Passing parameters
• Register stack engine

Crash Dump Fundamentals

• Overview of OpenVMS crashes
• Crash dump analysis tools
• General steps in analyzing crashes

Interrupts and Exceptions

• Interrupts and exceptions
• Software Interrupt Services (SWIS)
• Exception frames
• SWIS log

Exception-Related Crashes

• Exceptions and condition handling review
• Exception related crashes
• Examining unexpected system service exception crashes

Invalid Exception Crashes

• Invalid exception crash concepts
• Analyzing invalid exception crashes

Analysis of Page Fault with IPL too High Crashes

• Page fault with IPL too high crash concepts
• Analyzing page fault with IPL too high crashes

More on SDA

• SDA techniques
• Monitoring oriented parameters

Kernel Stack Not Valid Crashes

• Kernel stack not valid concepts
• Sample analysis of kernel stack not valid crashes

SMP Related Crashes

• Symmetric multiprocessing
• SMP bugchecks
• Spin wait crashes
• CPU sanity timeout crashes

AST Related Crashes

• AST concepts
• AST crash considerations
• Sample crashes in AST routines

System Hangs

• System hangs
• Methods for forcing a system crash
• Analyzing system hangs
• Sample analysis of system hangs