Re: Are there kernel testing suites out there? We need them.

Miles Lane (miles@amazon.com)
Sun, 04 Jul 1999 12:11:07 -0700


Dan Kegel wrote:
> The other main point about automated regression tests is that
> they take no work on the part of the kernel developers to run.

It would be great to have a Linux "Build Verification Test"
(pronounced "bee vee tee" in the testing world) suite. The
suite verifies that all essencial system functions are working
within the bounds of normal, correct usage. BVTs could be
written on a subsystem by subsystem basis. The Test Script
Manager would check to see what hardware was detected at boot
time by examining "dmesg" (if Linux ever gets a full
implementation of Plug and Play, the script could perform
PNP detection). Then the Test Script Manager could fire off
tests as appropriate. The Script Manager should support
selecting various test suite profiles:

Build Verification Testing
Stress Testing
Application Testing (we'd need some sort of
test harness that would allow scripted
driving of applications through both X
and command shells)
And so on...

> At the moment, the regression test for new Linux kernels
> is "Let a couple hundred people download it and see
> if they notice any problems" :-)

Woo hoo! The thrill...

> That works, but adding a mechanized regression test would
> be painless, and might give us earlier warnings if and when
> a few bad things (like ext2 fs corruption) creep back into
> the kernel. Earlier warnings about problems are what
> corporations love. Red Hat, for instance, might well feel
> inclined to set up a mechanized regression test to give
> it a little added certainty that it's not about to ship a lemon.
> - Dan

Well said. And much more succinctly than I. :-)

Miles

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/