ZJDBGPack 2.0 improved “batch-ability”
An improved version of ZJDBGPack has been released, with better error messages and non-zero exit codes when an error occurs. This makes it more usable for batches and automated builds.
An improved version of ZJDBGPack has been released, with better error messages and non-zero exit codes when an error occurs. This makes it more usable for batches and automated builds.
SamplingProfiler has a few options to help profile a multi-threaded application which I’ll go over here.
In the current version, those options allow identifying CPU-related bottlenecks, as in “threads taking too much CPU resources or execution time”. However, they do not provide much clues yet to pinpoint bottlenecks arising from thread synchronization issues or serialization (insufficient parallelism). Hopefully, more support for profiling multi-threaded applications will come in future versions.
ZJDBGPack is again available, but as an independent download (it used to be bundled with SamplingProfiler).
This is a command-line utility intended for use in a build process or from the Delphi tools menu, whose purpose is to integrate debug information into an executable. The debug information format is a compressed version of JCL‘s JDBG.
As of know, SamplingProfiler is the only published utility that understands this format, so you can use it either to reduce the size of the executables you deploy for profiling purposes, or if you do not want to deploy directly-readable debug information files.
This morning while debugging a statistical ichthyo-parser I stumbled upon what looked like a Delphi 2009 compiler bug: the compiler was outputting gibberish ASM opcodes… But after further investigations, it appeared this wasn’t completely gibberish, but that it was (somewhat) correct MSIL bytecode!
One issue when trying to profile a “live” application is that you may be getting a lot of noise, resulting from a particular library or section of code being executed from multiple contexts.
You may also be after profiling only one particular case, and want some reproducibility between runs… in short: you want a finer grained control on when or for what the profiling will take place.