Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

NOTES checker

2 views
Skip to first unread message

John Baldwin

unread,
Dec 16, 2009, 5:22:49 PM12/16/09
to cur...@freebsd.org
I wrote a little script today that crawls through sys/conf/files*,
sys/conf/options*, sys/conf/NOTES, and sys/*/conf/NOTES looking for devices
or options that are not tested in all the places they are valid. If folks
are interested I can commit it to src/tools. At the moment we have 300+
violations in HEAD. :) The first few look like this:

WARN: Attempt to change type of gdc from option to device at /home/jhb/work/freebsd/svn/head/sys/pc98/conf/NOTES:307
Previous type set at /home/jhb/work/freebsd/svn/head/sys/conf/options.pc98:70
WARN: device acpi_aiboost is defined globally but only tested in i386 NOTES
WARN: device acpi_asus is defined globally but only tested in i386 NOTES
WARN: device acpi_dock is defined globally but only tested in i386 NOTES
WARN: device acpi_fujitsu is defined globally but only tested in i386 NOTES
WARN: device acpi_hp is defined globally but only tested in i386 NOTES
WARN: device acpi_ibm is defined globally but only tested in i386 NOTES
WARN: option ACPI_MAX_TASKS is defined globally but never tested
WARN: option ACPI_MAX_THREADS is defined globally but never tested
WARN: device acpi_panasonic is defined globally but only tested in i386 NOTES
WARN: device acpi_sony is defined globally but only tested in i386 NOTES
WARN: device acpi_toshiba is defined globally but only tested in i386 NOTES
WARN: device acpi_video is defined globally but only tested in i386 NOTES
WARN: device acpi_wmi is defined globally but only tested in i386 NOTES
WARN: ad7418 is defined globally but never tested
WARN: ada is defined globally but never tested
WARN: option ADAPTIVE_LOCKMGRS is defined globally but never tested
WARN: adlink is defined globally but never tested

--
John Baldwin
_______________________________________________
freebsd...@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-current
To unsubscribe, send any mail to "freebsd-curre...@freebsd.org"

Ollivier Robert

unread,
Dec 16, 2009, 5:48:23 PM12/16/09
to freebsd...@freebsd.org
According to John Baldwin:

> I wrote a little script today that crawls through sys/conf/files*,
> sys/conf/options*, sys/conf/NOTES, and sys/*/conf/NOTES looking for devices
> or options that are not tested in all the places they are valid. If folks
> are interested I can commit it to src/tools. At the moment we have 300+
> violations in HEAD. :) The first few look like this:

Sounds like a perfect candidate for src/tools or regression tests.

--
Ollivier ROBERT -=- FreeBSD: The Power to Serve! -=- rob...@keltia.freenix.fr
In memoriam to Ondine : http://ondine.keltia.net/

Ivan Voras

unread,
Dec 17, 2009, 12:49:12 PM12/17/09
to freebsd...@freebsd.org
John Baldwin wrote:
> I wrote a little script today that crawls through sys/conf/files*,
> sys/conf/options*, sys/conf/NOTES, and sys/*/conf/NOTES looking for devices
> or options that are not tested in all the places they are valid. If folks
> are interested I can commit it to src/tools. At the moment we have 300+
> violations in HEAD. :) The first few look like this:

What does "tested" mean in this context? I'm pretty sure some of the
devices/drivers in the list work but that's probably not it :)

>
> WARN: Attempt to change type of gdc from option to device at /home/jhb/work/freebsd/svn/head/sys/pc98/conf/NOTES:307
> Previous type set at /home/jhb/work/freebsd/svn/head/sys/conf/options.pc98:70
> WARN: device acpi_aiboost is defined globally but only tested in i386 NOTES
> WARN: device acpi_asus is defined globally but only tested in i386 NOTES
> WARN: device acpi_dock is defined globally but only tested in i386 NOTES
> WARN: device acpi_fujitsu is defined globally but only tested in i386 NOTES
> WARN: device acpi_hp is defined globally but only tested in i386 NOTES
> WARN: device acpi_ibm is defined globally but only tested in i386 NOTES
> WARN: option ACPI_MAX_TASKS is defined globally but never tested
> WARN: option ACPI_MAX_THREADS is defined globally but never tested
> WARN: device acpi_panasonic is defined globally but only tested in i386 NOTES
> WARN: device acpi_sony is defined globally but only tested in i386 NOTES
> WARN: device acpi_toshiba is defined globally but only tested in i386 NOTES
> WARN: device acpi_video is defined globally but only tested in i386 NOTES
> WARN: device acpi_wmi is defined globally but only tested in i386 NOTES
> WARN: ad7418 is defined globally but never tested
> WARN: ada is defined globally but never tested
> WARN: option ADAPTIVE_LOCKMGRS is defined globally but never tested
> WARN: adlink is defined globally but never tested
>

_______________________________________________

Dag-Erling Smørgrav

unread,
Dec 18, 2009, 7:28:19 AM12/18/09
to Ivan Voras, freebsd...@freebsd.org
Ivan Voras <ivo...@freebsd.org> writes:

> John Baldwin <j...@freebsd.org> writes:
> > I wrote a little script today that crawls through sys/conf/files*,
> > sys/conf/options*, sys/conf/NOTES, and sys/*/conf/NOTES looking for devices
> > or options that are not tested in all the places they are valid. If folks
> > are interested I can commit it to src/tools. At the moment we have 300+
> > violations in HEAD. :) The first few look like this:
> What does "tested" mean in this context?

IIUC, it means "listed in NOTES on all arches that it claims to run on".

> > WARN: device acpi_aiboost is defined globally but only tested in i386 NOTES
> > WARN: device acpi_asus is defined globally but only tested in i386 NOTES
> > WARN: device acpi_dock is defined globally but only tested in i386 NOTES
> > WARN: device acpi_fujitsu is defined globally but only tested in i386 NOTES
> > WARN: device acpi_hp is defined globally but only tested in i386 NOTES
> > WARN: device acpi_ibm is defined globally but only tested in i386 NOTES
> > WARN: option ACPI_MAX_TASKS is defined globally but never tested
> > WARN: option ACPI_MAX_THREADS is defined globally but never tested
> > WARN: device acpi_panasonic is defined globally but only tested in i386 NOTES
> > WARN: device acpi_sony is defined globally but only tested in i386 NOTES
> > WARN: device acpi_toshiba is defined globally but only tested in i386 NOTES
> > WARN: device acpi_video is defined globally but only tested in i386 NOTES
> > WARN: device acpi_wmi is defined globally but only tested in i386 NOTES

The bug here is not that they aren't tested outside i386, but that they
are defined globally (although some of them, perhaps even all, are also
valid on amd64)

DES
--
Dag-Erling Smørgrav - d...@des.no

John Baldwin

unread,
Dec 21, 2009, 8:39:32 AM12/21/09
to freebsd...@freebsd.org, Dag-Erling Smørgrav, Ivan Voras
On Friday 18 December 2009 7:27:20 am Dag-Erling Smørgrav wrote:
> Ivan Voras <ivo...@freebsd.org> writes:
> > John Baldwin <j...@freebsd.org> writes:
> > > I wrote a little script today that crawls through sys/conf/files*,
> > > sys/conf/options*, sys/conf/NOTES, and sys/*/conf/NOTES looking for devices
> > > or options that are not tested in all the places they are valid. If folks
> > > are interested I can commit it to src/tools. At the moment we have 300+
> > > violations in HEAD. :) The first few look like this:
> > What does "tested" mean in this context?
>
> IIUC, it means "listed in NOTES on all arches that it claims to run on".

Yes. It means something like 'LINT builds will test this option'.

> > > WARN: device acpi_aiboost is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_asus is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_dock is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_fujitsu is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_hp is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_ibm is defined globally but only tested in i386 NOTES
> > > WARN: option ACPI_MAX_TASKS is defined globally but never tested
> > > WARN: option ACPI_MAX_THREADS is defined globally but never tested
> > > WARN: device acpi_panasonic is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_sony is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_toshiba is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_video is defined globally but only tested in i386 NOTES
> > > WARN: device acpi_wmi is defined globally but only tested in i386 NOTES
>
> The bug here is not that they aren't tested outside i386, but that they
> are defined globally (although some of them, perhaps even all, are also
> valid on amd64)

Yep. Either way it is still "buggy".

--
John Baldwin

0 new messages