firmware-testing-team team mailing list archive

[Colin Ian King <colin.king@xxxxxxxxxxxxx>]

Hi once again,

lots to report on, it has been a busy month!

== News ==

Ubuntu Natty 11.04 feature freeze means that we reach a point where we
drawn a line for any further changes for fwts for the Natty cycle.
Version 0.22.03 is now in the Ubuntu Natty Universe archive and from now
on any changes for Natty will be just bug fixes.

The fwts PPAs have been re-organised (again) since my last message, so
apologies for another change, hopefully it will be the last.

I've organised the PPAs as follows:

        Stable: for the latest stable code, e.g. stable code + bug fixes
        that are currently in Natty
        
        Devel: development code, for the new 'crack-of-the-day'
        features. Expect new features and maybe bugs.

..both these will contain packages built for the Maverick and Natty
series, and future series too.  Thus Maverick users can grab the latest
stable code found in Natty by subscribing to the Stable PPA

Installation:

For Stable fwts:
        sudo add-apt-repository
        ppa:firmware-testing-team/ppa-fwts-stable
        sudo apt-get update
        sudo apt-get install fwts
        
For Development fwts:
        sudo add-apt-repository ppa:firmware-testing-team/ppa-fwts-devel
        sudo apt-get update
        sudo apt-get install fwts
        
Next month I will be deleting the older PPAs as they are now redundant.
        
== So what has landed since last month? ==

Actually, quite a lot between 0.21.04 and 0.22.03. Most of these changes
are internal changes to fwts, a bunch of bug fixes and some extra
features for s3/s4 support.

Internal fwts changes:
 * make option handling more extensible and maintainable.
 * make fwts core library and ACPICA libraries into share libraries.  
 * use less dynamically allocated list headers where appropriate.
 * fix build ordering and lintian warnings.

Bug fixes:
 * bunch of typos in code and man page.
 * memory leaks.
 * dmar test, mark test as skipped if DMAR non-existent.
 * smbios test, return pass/failed status.
 * virt test, machines without virt support now marked as skipped.
 * crs test, mark untested runs as skipped.
 * apicinstance: mark test as passed if it is correct.
 * dmar test, don't double free table.

Feature changes:
 * remove deprecated options from code and man page.
 * add ~20 more method/object checks in method test.
 * summary table now fit into 80 columns.
 * collate summary data into a more compact form.
 * improve progress feedback when scanning huge logs.
 * add _BCM execution failure check.
 * add more intelligence to FADT checking.
 * s3/s4: add --s3-device-check, --s4-device-check.
   to check if devices re-configure correctly over cycles.
 * s3/s4: add --s4-device-check-delay, --s4-device-check-delay
   to wait for devices to settle before running device check.
 * s3/s4: add --s3-quirks, --s4-quirks to pass quirking over to
   pm-suspend, pm-hibernate.
 * s3/s4: remove two phases of testing, roll into one. This now
   provides sane progress feedback and rationalises the code. 
 * s3/s4: collect pass/fail results to end of tests rather than
   reporting these per cycle. This reduces log bloat when doing
   extended soak testing.
 * s3/s4: more progress feedback during waits.
 * mtrr test: disable MTRR VGA test - it is an unhelpful test.


== What is Next? ==

I've now created a o-release branch in the fwts git repo for newer
features that will land in the Natty+1 release and already I've added
one more utility which did not quite make it for Natty.  I will change
the name of the o-release branch once we know what Natty+1 will be
called.

New code from this o-release branch will now be ending up in the
Development PPA. I will bump the version number to 0.23.00 for this new
development cycle.

The new o-release branch contains a utility "mpdump" which looks for the
Intel MultiProcessor MP table and will dump out the contents. This
contains the BIOS interrupt configuration and some system memory maps.
This will hopefully help us eyeball problematic APIC configurations, for
example, broken polarity settings.

== Thanks to.. ==

I'd like to thank Kamal Mostafa for his help and time in cleaning up the
lintian warnings and fixing the somewhat poorly organised link ordering.
Also, thanks to
Seth Forshee for picking up a bug in the HPET results when no base
address is found and also contributing a build fix.  I appreciate all
the feedback this month during some heavy soak testing with fwts, this
has helped us add some more features to s3/s4 testing.

I really look forward to your contributions, be it bug fixes, extra
tests, bug reports and requests for extra features. Keep them coming!
And feel free to discuss these in this mailing list.

Colin