Re: Regression tests, benchmark histories (Was: (reiserfs, ext2 resizing patches, etc.))

From: Jeff Garzik (
Date: Tue Jun 20 2000 - 10:16:23 EST

Larry McVoy wrote:
> > Would it not be helpful to create some regression tests for core
> > subsystems like ext2? Then you could refactor it at will: as long as it
> > still runs the regression tests, you know the change didn't break
> > anything. (Of course, if the change does break something, the
> > regression tests need to be extended, not condemned.)
> We have this sort of thing for BitKeeper, we're 12009 lines of shell, and
> it is a huge win.
> The smart thing to do is to make it easy to add a test and to run the test.
> We have the infrastructure set up such that you can create tests easily,
> this is a self contained test:

If anyone is volunteering to setup and maintain this sort of thing, I've
been collecting test programs of various sorts and I'll bet lkml has
tons of suggestions.

Unfortunately regression tests in particular aren't plentiful because a
lot of bug reports are handled without specifically creating a test case
which exposes a problem. Also recognize that there are a class of
problem for which creating tests is difficult ("it won't boot when XXX
switch is flippped in BIOS").


Jeff Garzik              |
Building 1024            | Free beer tomorrow.
MandrakeSoft, Inc.       |

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to Please read the FAQ at

This archive was generated by hypermail 2b29 : Fri Jun 23 2000 - 21:00:20 EST