Wei Liu [Mon, 2 Feb 2015 19:57:13 +0000 (19:57 +0000)]
mfi-common, make-flight: create XSM test jobs
Duplicate Debian PV and HVM test jobs for XSM testing.
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: Ian Jackson <ian.jackson@eu.citrix.com>
---
Changes in v8:
1. Make libvirtbuildjob = ${bfi}build-$dom0arch-libvirt
Changes in v6:
1. Skip generating xsm job for different platforms.
2. Use "xsms".
3. Reformat some long lines.
Wei Liu [Mon, 2 Feb 2015 19:53:26 +0000 (19:53 +0000)]
make-flight: factor out do_pv_debian_tests
Pure code motion. No effect on job generation.
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Wei Liu [Fri, 12 Sep 2014 15:29:00 +0000 (16:29 +0100)]
Debian.pm: load flask policy in uboot
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: Ian Jackson <ian.jackson@eu.citrix.com>
---
Changes in v10:
1. Correctly get $flaskpolicy.
Changes in v9:
1. Add "xen,multiboot-module".
Changes in v8:
1. Append flask_enforcing=1 and flask_enabled=1.
Wei Liu [Mon, 8 Sep 2014 10:41:44 +0000 (11:41 +0100)]
mfi-common: create build-$arch-xsm job
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
---
Changes in v4:
1. Use "true" and "false" instead of "y" and "n".
2. Rename xenbranch_wants_xsm_tests to xenbranch_xsm_variants.
Wei Liu [Mon, 8 Sep 2014 15:06:52 +0000 (16:06 +0100)]
ts-xen-build: build with XSM support if requested
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Cc: Ian Jackson <ian.jackson@eu.citrix.com> Cc: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
---
Changes in v5:
1. Only set XSM_ENABLE when runvar is defined.
2. Fix inconsistent whitespace.
Wei Liu [Sun, 12 Oct 2014 16:04:34 +0000 (17:04 +0100)]
overlay: update overlay/etc/grub.d/20_linux_xen
This file was originally created to work around Debian bug #633127
("/etc/grub/20_linux does not recognise some old Xen kernels").
According to Debian bug tracker [0], #633127 bug is fixed in Wheezy. As
we're now using Wheezy in OSSTest we can safely remove the old overlay
file if there's no further bugs discovered.
However we have another bug #690538 ("grub-common: Please make submenu
creation optional or at least allow users to disable it easily") that
would break OSSTest. We're now using Wheezy in production. There's no
way to disable submenu in Wheezy. And submenu breaks OSSTest's grub menu
parser.
So update this overlay file to the one in Wheezy's grub-common
1.99-27+deb7u2 and take care of Debian bug #690538 by removing the lines
to generate submenu.
Also work around GRUB bug #43420 ("20_linux_xen doesn't support Xen XSM
policy file") by applying a small patch proposed in [2].
Add a note to reference #633127 and #690538 above grub2 setup function.
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Cc: Ian Jackson <ian.jackson@eu.citrix.com> Cc: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Tue, 17 Feb 2015 17:35:57 +0000 (17:35 +0000)]
NTP servers: Work around Debian's failure to honour preseed
Setting clock-setup/ntp-server is not sufficient: it only takes effect
in the installer (!)
I have reported this as Debian #778564. In the meantime we should
work around it for current releases (including jessie, which is
frozen).
For later releases, the new ntp.conf editing code arranges to bomb out
if we have an NTP server configured and find it hasn't been honoured
during the install.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
-### END OF DEBIAN PRESEED BASE
+END
+
+ my $ntpserver = get_target_property($ho,'NtpServer');
+use Data::Dumper;
+print STDERR "PRESEED NTP ", Dumper($ntpserver);
+ $preseed .= <<"END" if $ntpserver;
+d-i clock-setup/ntp-server string $ntpserver
+END
+ $preseed .= <<"END";
+
+### END OF DEBIAN PRESEED BASE
END
-}
+
+ return $preseed;
+}
sub preseed_create ($$;@) {
my ($ho, $sfx, %xopts) = @_;
diff --git a/README b/README
index 6e63e97..0a0242c 100644
--- a/README
+++ b/README
@@ -334,6 +334,11 @@ HostProp_<testbox>_TftpScope
Defines the Tftp scope (i.e. subnet) where this host resides. See
"TftpFoo_<scope> and TftpFoo" below.
+HostProp_<testbox>_NtpServer
+ NTP server to use. You should probably have your own local
+ NTP server for production use; the default is to use the operating
+ system's default (normally, Debian's pool.ntp.org servers).
+
HostFlags_<testbox>
Defines a set of flags for the host. Flags is a list separated by
whitespace, comma or semi-colon. A flag can be unset by prepending
@@ -357,9 +362,7 @@ HostGroupFlags_<group>
merged with the host specific flags. Only used in standalone mode.
DebianPreseed
- Text to add to the debian-installer preseed file. Optional
- but you will need to set some NTP servers here if your firewall
- doesn't permit NTP to Debian's pool.ntp.org servers.
+ Text to add to the debian-installer preseed file. Optional.
-### END OF DEBIAN PRESEED BASE
+END
+
+ my $ntpserver = get_target_property($ho,'NtpServer');
+use Data::Dumper;
+print STDERR "PRESEED NTP ", Dumper($ntpserver);
+ $preseed .= <<"END" if $ntpserver;
+d-i clock-setup/ntp-server string $ntpserver
+END
+ $preseed .= <<"END";
+
+### END OF DEBIAN PRESEED BASE
END
-}
+
+ return $preseed;
+}
sub preseed_create ($$;@) {
my ($ho, $sfx, %xopts) = @_;
diff --git a/README b/README
index 6e63e97..0a0242c 100644
--- a/README
+++ b/README
@@ -334,6 +334,11 @@ HostProp_<testbox>_TftpScope
Defines the Tftp scope (i.e. subnet) where this host resides. See
"TftpFoo_<scope> and TftpFoo" below.
+HostProp_<testbox>_NtpServer
+ NTP server to use. You should probably have your own local
+ NTP server for production use; the default is to use the operating
+ system's default (normally, Debian's pool.ntp.org servers).
+
HostFlags_<testbox>
Defines a set of flags for the host. Flags is a list separated by
whitespace, comma or semi-colon. A flag can be unset by prepending
@@ -357,9 +362,7 @@ HostGroupFlags_<group>
merged with the host specific flags. Only used in standalone mode.
DebianPreseed
- Text to add to the debian-installer preseed file. Optional
- but you will need to set some NTP servers here if your firewall
- doesn't permit NTP to Debian's pool.ntp.org servers.
+ Text to add to the debian-installer preseed file. Optional.
Ian Jackson [Tue, 17 Feb 2015 17:21:47 +0000 (17:21 +0000)]
TestSupport: Provide get_target_property
This looks in the supplied $ho, but if that's a $gho (ie it has a
$gho->{Host}) it also looks in its host.
This is going to be useful for a fair variety of host-specific or
infrastructure-determined properties.
It seems to me that whether a property ought to be looked up in the
host if not found in a particular guest depends mostly on the
property, and not on the way the guest is configured. The easiest way
to represent that in the osstest codebase is probably to call
get_target_property instead of get_host_property in the appropriate
places.
The use of recursion will make this look through a series of nested
hosts if we have nested virtualisation going on. Indeed, nested
virtualisation may benefit from replacement of get_host_property by
get_target_property in a number of cases. (At the time of writing
there is no nested virt in osstest mainline, but it's on the way.)
Currently there are no callers of get_target_property. One will
appear shortly.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com> CC: Robert Hu <robert.hu@intel.com> CC: LongtaoX Pang <longtaox.pang@intel.com>
Ian Jackson [Fri, 6 Feb 2015 17:09:40 +0000 (17:09 +0000)]
rump kernel tests: Repeat the xenstorels test 50 times
Add a new step which uses repeat-ts to run
ts-rumpuserxen-demo-xenstorels many times.
We have to run ts-guest-destroy-hard after each time, to destroy the
guest which the demo script leaves lying about.
Strategically placed `+'s in the repeat-ts command line arrange that
the testid ends up being
rumpuserxen-demo-xenstorels/xenstorels.repeat
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
---
v2: Run the test after, rather than before, the explicit
ts-guest-destroy-hard. That will avoid blocking the single
destroy test if the repeat fails.
No longer specify to tolerate failures of the post-run-demo
destroy, as if the test passes so must the destroy. Now by-hand
testing may need a different ts-repeat-test rune, but in practice
by-hand testing will probably involve a shell loop or something
anyway.
Ian Jackson [Fri, 6 Feb 2015 17:08:31 +0000 (17:08 +0000)]
rump kernel tests: Cancel unneeded edits of guest config
If the guest config is already set up to preserve, cancel the edit.
We are going to repeat this test, and this avoids creating many
identical copies of the same file in the log output.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbelL@citrix.com>
Ian Jackson [Fri, 6 Feb 2015 15:36:34 +0000 (15:36 +0000)]
sg-run-job: testid generation: Process ts more like rest of args
Remove ts as a separate parameter to spawn-ts. The test script now
becomes the first entry in args.
We process it through the arg loop as before. Currently there are no
calls where the first arg is `+' so the test script name ends up in
both real_args and testid_args.
We split it out of real_args into the ts variable with lshift.
We split it out of testid_args into the deftestid with lshift.
So afterwards in spawn-ts, all the variables (including real_args, ts,
deftestid and testid_args and hence host_testid_suffix) have the
values they would have had before.
Therefore there is no functional change for any existing calls.
However, because the first argument is not treated specially for the
`+' procesing loop, it is now possible to specify `+' as the first
entry in args to spawn-ts (ie where ts used to be) to arrange that the
deftestid (and hence, probably, the testid) is computed using later
arguments.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Ian Jackson [Mon, 9 Feb 2015 17:55:05 +0000 (17:55 +0000)]
sg-report-job-history: Show which host test ran on
Add a column listing the host(s) used. We first find the relevant set
of host runvars (with a SELECT DISTINCT) and then look up each var for
each actual flight.
We do the pattern-matching on runvar names in perl to avoid giving the
postgresql optimiser a chance to turn this query into a full table
scan of the runvars table. (A previous iteration of this patch
searched, in the per-flight loop, for all runvars whose name was
appropriately LIKE, with appalling performance.)
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Mon, 9 Feb 2015 14:22:50 +0000 (14:22 +0000)]
allow.all: Do not regard libvirt guest start failures as regressions
We have been experiencing some regressions in test-amd64-*-libvirt
guest-start.
These appear to be host-specific: for example, test-amd64-i386-libvirt
succeeded in 34278 on itch-mite but failed in 34234 on bush-cricket.
The only difference in osstest's records of the versions is a change
to linuxfirmware.
I booked out field-cricket and have repro'd some problems there. It
appears that libvirtd can crash. The symptoms are not the same as in
the push gate failures in osstest (34316, 34234) and linux-linus
(34299) but similar to those in linux-3.0 (34291) and linux-3.16
(34285).
When running libvirtd in the foreground we can get this:
libvirtd: libxl_event.c:1791: libxl__ao_complete_check_progress_reports: Assertion `ao->in_initiator' failed.
So, it appears to be some kind of race. The libvirt tests are new so
this is not a real regression in any of the trees gated by osstest.
So, permit this test for now.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Ian Campbell [Fri, 30 Jan 2015 10:55:03 +0000 (10:55 +0000)]
README.dev: Updates to commissioning process
- mkpxedir requires the resource to be allocated
- mention the need to bless the hosts for comissioning run, and to add
proper blessings afterwards.
- run mg-execute-flight in screen
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Wed, 4 Feb 2015 09:49:44 +0000 (09:49 +0000)]
ts-logs-capture: ensure captured log files are readable.
At least /var/log/libvirt/libvirtd.log is 0600 on the target machine
and this gets preserved in the captured logfile, eventually meaning
that apache on the log server cannot read the file to serve it ("You
don't have permission to access...").
Set all logfiles to 0666, modified by umask.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Wei Liu [Sat, 31 Jan 2015 22:51:50 +0000 (22:51 +0000)]
ts-xen-build-prep: install nasm
OVMF requires nasm to build.
Signed-off-by: Wei Liu <wei.liu2@citrix.com> Cc: Ian Campbell <ian.campbell@citrix.com> Cc: Ian Jackson <ian.jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Tue, 2 Dec 2014 12:53:26 +0000 (12:53 +0000)]
ts-guest-start: Use guest_create
This allows us to abolish CfgPathVar which was inconsistently used,
appears redundant with $gho->{CfgPath} and in any case never set to
anything other than 'cfgpath'.
This was intended to deal with toolstacks with a different cfg format
(such as very early versions of xl) but is no longer relevant.
I think if this need arises again in a future toolstack this
functionality could be reintroduced pretty trivially via the toolstack
abstraction which is added by this series.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Fri, 28 Nov 2014 17:35:09 +0000 (17:35 +0000)]
Toolstack: Refactor migration support.
Note that since the previous patch arranges for
ts-migration-support-check to continue to fail for libvirt the libvirt
code is not actually called yet (and will die if it is). This patch is
mainly included to reduce the number of users of
toolstack()->{Command} closer to zero.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Mon, 16 Jun 2014 15:46:40 +0000 (16:46 +0100)]
Toolstack: Refactor guest lifecycle.
Implement destory/create as per toolstack methods, including
implementing the libvirt version which previously didn't work. To do
this we use the virsh capability to convert an xl/xm style config file
into the correct XML.
xend basically calls into the xl helper since they are compatible.
xl/xm, uses ->{Command} which will eventually become private.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Mon, 16 Jun 2014 12:29:13 +0000 (13:29 +0100)]
TestSupport: always use xl for generic operations.
Unless the toolstack is xend (for compatibility with pre-xl Xen
versions), when we use xm.
For several operations in TestSupport.pm the actual toolstack isn't
really relevant, since we want info straight from Xen. For simplicity
just use xl (or xm) in these cases, to avoid needing to implement the
following specially for each toolstack:
- host_get_free_memory
- guest_get_state
- guest_find_domid
- listing assignable pci devices
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Mon, 16 Jun 2014 08:18:35 +0000 (09:18 +0100)]
Toolstack: use get_host_method_object() to manage toolstack selection
This will allow us to more easily have per-toolstack methods etc.
The previous hash of toolstack parameters is now a blessed object. For
now the callers don't need to change but over the following patches we
will refactor things to use method calls. In particular we will be
aiming to remove Command from the hash and use method calls for
everything.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Mon, 16 Jun 2014 12:00:14 +0000 (13:00 +0100)]
Pass host to toolstack()
This will be needed in a future patch.
Everywhere already has a $ho in hand. Also cache the answer as
$ho->{Toolstack}.
I scanned the source with:
find -name \*.pm -exec perl -c {} \;
for i in ts-* ; do perl -c $i; done
which reported "Not enough arguments for Osstest::TestSupport::toolstack"
for each callsite which needed changing.
Also don't pass the toolstack command name directly to
host_get_free_memory(). Look it up instead.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Tue, 15 Apr 2014 10:39:32 +0000 (11:39 +0100)]
apt: lock osstest's usages of apt-get against each other
Currently we rely on all apt-get invocations being in a single
ts-xen-build-prep job which can't run on a shared host.
That is a bit inflexible so instead use our own lock. We wait
indefinitely and rely on osstest's existing command timeout
infrastructure to catch problems.
target_install_packages*() previous estimated the time taken to
install the packages based on the number of packages. This no longer
applies because the install might get stuck behind some other large
install. Use a 3000s (nearly an hour) timeout instead (I expect
failures here to be unusual so erred on the big side)
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Jackson [Thu, 29 Jan 2015 12:49:48 +0000 (12:49 +0000)]
ts-hosts-allocate-Executive: Fix an undefined warning
$esrow is autovivified as { } by
$candrow->{EquivMostRecentStatus} = $esrow->{status}
so use %$esrow which in scalar context is true iff the
hash is nonempty.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
Ian Jackson [Mon, 26 Jan 2015 18:05:11 +0000 (18:05 +0000)]
linux-3.16: New branch
Test this too. We fetch this from the Canonical Kernel Team (ckt)
repo, but do not encode `ckt' in the branch name, in case this becomes
a proper stable tree.
(I have set up this branch on the osstest with the mg-branch-setup
script in this version of this tree.)
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
Dario Faggioli [Thu, 8 Jan 2015 17:47:02 +0000 (18:47 +0100)]
make-flight: reorganize scheduling related test jobs
Scheduling related tests are ok to run on ARM, so do
not cut them off. They also do not depend on a
particular Dom0 architecture.
The net effect is that the following tests are removed:
test-amd64-i386-xl-credit2
test-amd64-i386-xl-multivcpu
while the following new ones are created:
test-amd64-amd64-xl-credit2
test-amd64-amd64-xl-multivcpu
test-armhf-armhf-xl-credit2
test-armhf-armhf-xl-multivcpu
test-armhf-armhf-xl-sedf
test-armhf-armhf-xl-sedf-pin
Signed-off-by: Dario Faggioli <dario.faggioli@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Thu, 22 Jan 2015 15:53:35 +0000 (15:53 +0000)]
Switch ARM testing back to linux-arm-xen branch.
The cubietruck and arndale boards need a newer kernel than our current
baseline of 3.14. Therefore switch all ARM testing to the
linux-arm-kernel branch which has been moved up to test v3.16.
This effectively reverts commit 38b2fe2020883 "Switch armhf to use
common default Linux."
Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Ian Campbell [Thu, 22 Jan 2015 11:05:05 +0000 (11:05 +0000)]
Debian: Create boot.scr with a suffix and copy to boot.scr
This ensures that we always have a boot script to boot at least the
native Debian kernel and Xen available, even after multiple iterations
of installation, which is handy when debugging a system.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Thu, 22 Jan 2015 11:05:03 +0000 (11:05 +0000)]
Osstest/Debian: Add 0x prefix to $filesize
$filesize is an unprefixed hex number, but fdt set requires the 0x to
interpret it properly. See
http://lists.denx.de/pipermail/u-boot/2014-October/193622.html,
http://lists.denx.de/pipermail/u-boot/2014-November/194150.html and
http://lists.denx.de/pipermail/u-boot/2014-November/194150.html.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Thu, 22 Jan 2015 11:05:02 +0000 (11:05 +0000)]
Osstest/Debian: Add "clk_ignore_unused" to default command line
dom0 is not aware that some clocks are actually in use (e.g. by the
hypervisor), so this stops the kernel from messing with (specifically,
disabling) those clocks. It's harmless even when not needed.
Really there ought to be some interface to communicate this from Xen
to dom0, or some other mechanism to gate things. See
http://bugs.xenproject.org/xen/bug/45
This bug is present in at least Wheezy and Jessie so apply this
workaround to those.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Thu, 22 Jan 2015 11:04:57 +0000 (11:04 +0000)]
Osstest/Debian: Support for loading an FDT from u-boot script
The currently supported platform provides an FDT preloaded at 0x1000.
Replace this with ${fdt_addr} (which the current platform exposes) and
for platforms which do not provide an fdt arrange to load the relevant
file as named in the ${fdtfile} (which is conventionally provided by
u-boot for platforms which need this).
Drop some random memory clearing rune, I've no idea what the intended
purpose was, 0x800000 doesn't correspond to any $foo_addr_r on the
midway systems for example.
Also get rid of the scsi scan which must necessarily have already
happened (since the boot.scr itselfs lives on a scsi drive).
Lastly, add some echos to show progress through the script, as an aid
to debugging.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Thu, 22 Jan 2015 11:04:54 +0000 (11:04 +0000)]
make-flight: Run a basic test on each arm platform
Unlike x86 there is enough variation in the ARM platforms that it is
worth having a basic test on each as part of a standard run. This
relies on each host having an appropriate platform-$platform host
flag.
The existing test-ARCH-ARCH-xl test is retained as a floating test,
while a new variant is added for each distinct platform present in the
hostdb which is tied to that platform type. The intention is that only
arm platforms will have platforms at first, although perhaps
platform-intel and platform-amd could be added in the future too.
There are currently no platform-* flags in the hostdb, so tested with
s/platform-/equiv-/ and:
( set -ex ;
source ./cri-getplatforms ;
blessing=real ;
export OSSTEST_CONFIG=production-config ;
for p in '' `getplatforms "armhf"` ; do
set +x ;
echo PLATFORM: $p ;
done
)
which prints:
PLATFORM:
PLATFORM: marilith
and with s/armhf/amd64/:
PLATFORM:
PLATFORM: rackservers-s40670
PLATFORM: r310-moth
PLATFORM: rackservers-s40680
PLATFORM: dell-r310
PLATFORM: rackservers-s40679
PLATFORM: rackservers-s40663
PLATFORM: rackservers-q21011
Also tested in standalone mode with a ~/.xen-osstest/config
containing:
PlatformsArmhf midway cubietruck arndale
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Sat, 10 Jan 2015 14:58:02 +0000 (14:58 +0000)]
mg-debian-installer-update: produce deterministic output
Currently rerunning mg-debian-install-update when the external files
have changed still produces differences in the local files produced
during post-processing.
Avoid these differences by:
- Using gzip -n, which avoids storing a timestamp in the gzip
header (as well as the name, which we don't need).
- Using pax -M norm, which normalises all timestamps (among other
things, such as the owner, which we don't care about)
- Using tar --mtime, with a reference within the dpkg-deb created
hierarchy (which has timestamps from the package and is therefore
dependent only on the downloaded package revision)
With this the results of two invocations of
mg-debian-installer-update(-all) are identical (assuming no changes to
the downloaded files) as demonstrated by runnign this quick hack:
Ian Campbell [Mon, 5 Jan 2015 12:57:08 +0000 (12:57 +0000)]
ts-libvirt-build: use Osstest::BuildSupport::submodulefixup
Instead of cloning gnulib manually which can break if upstream gnulib
gets ahead of libvirt.git (which applies patches on the fly etc). By
using submodulefixup we automatically DTRT and use the version of
gnulib specified by the libvirt.git submodule metadata, but with a
runvar override if necessary.
This also removes a whole bunch of faffing in ap-*, cr-daily-branch
and mfi-common to get the version of gnulib to use, which was always a
bit of a wart (ungated for one thing...).
We continue to use --no-git and GNULIB_SRCDIR because otherwise
autogen.sh (via bootstrap) will force its own version, overwriting
what submodulefixup has done. For this we need a way to get the hash
representing the module, so introduce submodule_find (and rework
submodule_have in terms of it).
Tested in standalone mode with build-amd64-libvirt and
build-amd64-rumpuserxen (because I touched submodule_have, AFAICT the
bodges were not run). The libvirt build was tested both with the
automatic revisions and with:
revision_libvirt=2360fe5d24175835d3f5fd1c7e8e6e13addab629
revision_libvirt_gnulib=16518d9ed8f25d3e53931dd1aa343072933e4604
(used in successful libvirt flight 32648), in both cases confirming
that the build used the desired versions.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@citrix.com>
---
v2: Honour revision_libvirt_gnulib.
v3: Fix submodule_have, defined(&sub) is always true because defined
is special wrt &sub.
Ian Campbell [Thu, 4 Dec 2014 09:58:34 +0000 (09:58 +0000)]
Add basic PVH flights.
These are the usual PV debian flights with pvh=1 added to the
configuration file.
A job is created for each of Intel and AMD, although obviously AMD is
expected to fail at the moment.
In my testing I got:
(XEN) Attempt to create a PVH guest on a system without necessary hardware support
because my test box happens to be AMD.
I have confirmed that the pvh=1 option is correctly present in the
guest cfg for the new pvh job, and that no pvh= is present at all in
the existing test-amd64-amd64-xl job (which is expected and desired if
no pvh runvar is present).
Beyond that I've not tested this at all I fully expect even Intel to
fail in the first instance, due to issues such as lack of necessary
kernel options etc. I suggest to take this now and iterate on any
further changes.
Ian Campbell [Fri, 21 Nov 2014 13:16:58 +0000 (13:16 +0000)]
Osstest/Debian: Add support for "ExtraInitramfsModules" host property
The arndale platform needs a bunch of clk, phy and regulator stuff in order to
access its root filesystem. However mkinitramfs is not (currently?) able to
figure this out and therefore doesn't include them in the initrd. See
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=762042
Add a new host prop which can list these required additional module and
arranges for a suitable /etc/initramfs-tools/modules to be created on install.
Using the new HostGroupProp syntax the required modules are:
Ian Jackson [Mon, 8 Dec 2014 12:10:24 +0000 (12:10 +0000)]
TestSupport: use timeout(1)
If a command we run times out, the machinery in cmd() will arrange
for the ts-* script to spot the timeout, and stop waiting for it.
However it is also necessary for the command we ran to die. It has a
copy of the owner daemon fd, so if it doesn't, our resources won't get
freed. In sufficiently exciting bugs, our allocation might continue
indefinitely, while a subprocess of ours hangs on after we are long
gone.
timeout(1) does not print a message when the process times out (!) So
we can't do away with the logic in cmd(). We set the timeout(1)
timeout to 30s more than our own timeout, so that cmd() will time
out first and print a message.
We could use alarm(1) as we do in Osstest/Serial/sympathy.pm but that
program isn't packaged and its unsophisticated approach is not really
appropriate for arbitrary nonconsenting programs.
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com>
Ian Campbell [Fri, 21 Nov 2014 13:16:52 +0000 (13:16 +0000)]
Add simple helper to update DI for all architectures.
Uses DebianNonfreeFirmware, even (especially) for production, so move
the README stanza out of standalone only section. The current default
matches what is in the current production versions of DI.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Campbell [Fri, 21 Nov 2014 13:16:51 +0000 (13:16 +0000)]
ts-kernel-build: enable CONFIG_IKCONFIG{_PROC}
This makes the kernel's .config available in /proc/config.gz and
embeds a copy which can be extracted with
linux/scripts/extract-ikconfig (which I've not tried, but have no
reason to doubt).
Having this around can be handy with an older osstest installed test
box and to confirm you've booted the kernel you think you have when
you are messing with .config options.
Signed-off-by: Ian Campbell <ian.campbell@citrix.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com>
Ian Jackson [Thu, 20 Nov 2014 13:48:56 +0000 (13:48 +0000)]
linux-next tests: Use correct branch for baseline
Make cr-daily-branch honour an environment or setting variable
EXTRA_SGR_ARGS. In branch-settings.linux-next set it appropriately to
arrange that the linux-next test reports consider linux-linus tests as
interesting as well as just linux-next ones.
(We already use a flight from linux-linus for selecting the baseline
linux version.)
Signed-off-by: Ian Jackson <Ian.Jackson@eu.citrix.com> Acked-by: Ian Campbell <ian.campbell@citrix.com>
projects / people / liuw / osstest.git / log