This blog forms part of the http://performanceGuru.com website which focuses on performance tuning UNIX-like Operating Systems (including Linux).
There seems to have been some recent confusion
that my comment announcing the availability of a patch against the 2.6.13 kernel,
constituted a production endorsement of Systemtap. In case anyone else is not familiar with how the Linux Development process works, here goes:
- a patch is generated against the most recently released mainline kernel from Linus's tree, such as the one I cited above.
- this is sent to the mailing lists for the areas of interest associated with the patch, such as the Systemtap lists in this case.
- the patch is peer reviewed by the members of that group, which happened in this case and a number of issues where uncovered.
- the patch is the given wider distribution on LKML where is reviewed by a much larger group of peers.
- after this process is complete, if there is a general appetite, the patch is submitted to Andrew Morton and enters his -mm tree.
- once Andrew has the patch merged and is happy that is should enter the mainline, it enters Linus's tree, where it will go through maybe upto four release candidates as the release stabilises and is tested by a much wider group of people.
- Once the patch is in Linus's tree it may then be selected for inclusion into vendor kernels, or perhaps backported into earlier kernels by the various vendors depending on its value to them.
Obviously, as the post I made pointed people to what was clearly an initial patch release and a request for comments, I assumed I did not need to point out that this was not 'production ready', after all I have recently pointed out in my blog
that Systemtap is in early development.
Hope this clear things up - please feel free to post comments if you have any particular questions.