Commit Graph

240 Commits

Author SHA1 Message Date
Jiri Denemark
bb6cedd208 cpu_x86: Ignore enabled features for input models in x86DecodeUseCandidate
While we don't want to aim for the shortest list of disabled features in
the baseline result (it would select a very old model), we want to do so
while looking at any of the input models for which we're trying to
compute a baseline CPU model. Given a set of input models, we always
want to take the least capable one of them (i.e., the one with shortest
list of disabled features) or a better model which is not one of the
input models.

So when considering an input model, we just check whether its list of
disabled features is shorter than the currently best one. When looking
at other models we check both enabled and disabled features while
penalizing disabled features as implemented by the previous patch.

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-05-06 17:33:47 +02:00
Jiri Denemark
48341b025a cpu_x86: Penalize disabled features when computing CPU model
For finding the best matching CPU model for a given set of features
while we don't know the CPU signature (i.e., when computing a baseline
CPU model) we've been using a "shortest list of features" heuristics.
This works well if new CPU models are supersets of older models, but
that's not always the case. As a result it may actually select a new CPU
model as a baseline while removing some features from it to make it
compatible with older models. This is in general worse than using an old
CPU model with a bunch of added features as a guest OS or apps may crash
when using features that were disabled.

On the other hand we don't want to end up with a very old model which
would guarantee no disabled features as it could stop a guest OS or apps
from using some features provided by the CPU because they would not
expect them on such an old CPU.

This patch changes the heuristics to something in between. Enabled and
disabled features are counted separately so that a CPU model requiring
some features to be disabled looks worse than a model with fewer
disabled features even if its complete list of features is longer. The
penalty given for each additional disabled feature gets bigger to make
longer list of disabled features look even worse.

https://bugzilla.redhat.com/show_bug.cgi?id=1851227

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-05-06 17:33:47 +02:00
Jiri Denemark
63d633b9a4 cputest: Add some real world baseline tests
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-05-06 17:33:46 +02:00
Jiri Denemark
3daa68e265 cputest: Give better names to baseline tests
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-05-06 17:33:46 +02:00
Jiri Denemark
6aff36019b cputest: Drop some old artificial baseline tests
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-05-06 17:33:46 +02:00
Jiri Denemark
d2e4d66be3 cpu_map: Disable cpu64-rhel* for host-model and baseline
These ancient RHEL-only CPU models should not really be used by any CPU
definition created by libvirt. We keep them just for backwards
compatibility with domains which might still be using them.

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-05-06 17:33:46 +02:00
Tim Wiederhake
50a40c8d52 cpu-data.py: Query hyperv enlightenments
Reporting hv-* properties properly requires hv to be enabled,
see qemu commit 071ce4b03b.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2022-01-28 10:55:03 +01:00
Tim Wiederhake
d1873e03b4 cpu-data.py: Automatically adjust command line for kcpuid
kcpuid does not have a "-1" flag.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2021-05-07 18:02:48 +02:00
Tim Wiederhake
95a89ed012 cpu-data.py: Parse kcpuid output
Linux 5.13 introduces "kcpuid", a tool similar to "cpuid", see
https://lore.kernel.org/lkml/1614928878-86075-1-git-send-email-feng.tang@intel.com/

Output formats of cpuid and kcpuid differ slightly. This adds support
for the latter.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2021-05-07 17:14:53 +02:00
Tim Wiederhake
5a175b2ddc cpu-data.py: Factor out cpuid parsing
Preparation for next patch.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2021-05-07 17:13:35 +02:00
Tim Wiederhake
5c17a7ba41 cpumap: Add support for ibrs CPU feature
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-03-01 20:09:46 +01:00
Jiri Denemark
4901314d0d cpu_map: Remove intel-pt from x86 CPU models
As explained in QEMU commit 4c257911dcc7c4189768e9651755c849ce9db4e8
intel-pt features should never be included in the CPU models as it was
not supported by KVM back then and even once it started to be supported,
users have to enable it by passing pt_mode=1 parameter to kvm_intel
module. The Icelake-* CPU models with intel-pt included were added to
QEMU 3.1.0 and removed right in the following 4.0.0 release (and even in
3.1.1 maintenance release).

In libvirt 6.10.0 I introduced 'removed' attribute for features included
in our CPU model definitions which we can use to drop intel-pt from
Icelake-* CPU models. Back then I explained we can safely do so only for
features which could never be enabled, which is not the case of intel-pt.

Theoretically, it could be possible to create an environment in which
QEMU would enable intel-pt without asking for it explicitly: it would
need to use a new enough kernel (not available at the time of QEMU
3.1.0) and pt_mode KVM parameter in combination with QEMU 3.1.0 running
a domain with q35 machine type and all that on a CPU which didn't really
exist at that time.

Migrating such domain to a host with newer SW stack including libvirt
with this patch applied would result in incompatible guest ABI (the
virtual CPU would lose intel-pt). However, QEMU changed its CPU models
unconditionally and thus migration would not work even without this
patch. That said, it is safe to follow QEMU and remove the feature from
Icelake-* CPU models in our cpu_map.

https://bugzilla.redhat.com/show_bug.cgi?id=1853972

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Tim Wiederhake <twiederh@redhat.com>
2021-01-26 15:44:50 +01:00
Tim Wiederhake
10b5e789c5 cpu-data: Pretend to always run on logical processor #0
The output of cpuid depends on the logical processor id the process
runs on, as reflected by the "local apic id" present in cpuid leaves
(eax=1,ebx=0), (eax=11,ebx=0), and (eax=11,ebx=1). This produces
arbitrary changes in the output files that complicate comparisons.

This patch masks the occurences of the local apic id with 0x00, so
that two consecutive runs of "./cpu-data.py gather" produce identical
results.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-08 18:07:00 +01:00
Jiri Denemark
51d1a2cacf cpu-gather: Rename the script as cpu-data.py
It is now doing way more than gathering the CPU data from a host as the
other scripts were merged in it.

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2021-01-08 11:00:45 +01:00
Jiri Denemark
07b7fd6a29 cpu-gather: Remove redundant "processor" from CPU data file names
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2021-01-08 11:00:45 +01:00
Tim Wiederhake
f0a5cf4b8a cpu_map: Define and enable Snowridge model
Due to missing pdpe1gb support in the host CPU data, the CPU is still
incorrectly detected as Westmere-IBRS for host capabilities because we
don't have the option to disable features included in the base model
there.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 23:23:41 +01:00
Tim Wiederhake
13db542cf3 cpu_map: Add support for split-lock-detect CPU feature
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 23:23:31 +01:00
Tim Wiederhake
e06dd56032 cpu_map: Add support for core-capability CPU feature
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 23:23:04 +01:00
Tim Wiederhake
8c5c660b99 cpu_map: Add support for fsrm CPU feature
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 23:22:49 +01:00
Tim Wiederhake
59a585fdb0 cputestdata: Add test data for Snowridge
It's obvious the CPU model detection provides strange results, which
will be fixed by adding a new Snowridge CPU model few patches later.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 23:20:03 +01:00
Tim Wiederhake
a42adc2714 cpu-gather: Merge cpu-cpuid.py
Old usage:
  cpu-cpuid.py diff FILE...
New usage:
  cpu-gather.py diff FILE...

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:13:22 +01:00
Tim Wiederhake
bd05de35ef cpu-gather: Factor out call to cpu-cpuid.py
This is a preparatory step to merge cpu-cpuid.py.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:11:12 +01:00
Tim Wiederhake
ba05774f57 cpu-gather: Use actions instead of flags for action argument
This allows for the functionality of cpu-cpuid.py script to be
integrated more naturally in a later patch.

Changes the way this script should be called:
  cpu-gather.py                   -> cpu-gather.py
  cpu-gather.py --gather          -> cpu-gather.py gather
  cpu-gather.py --parse           -> cpu-gather.py parse
  cpu-gather.py --gather --parse  -> cpu-gather.py full

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:11:07 +01:00
Tim Wiederhake
d90738bacc cpu-cpuid: Deduplicate register list
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:11:01 +01:00
Tim Wiederhake
21d097c4e1 cpu-cpuid: Merge checkFeature functions
Prepare to deduplicate the list of relevant registers for cpuid and
msr information.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:10:56 +01:00
Tim Wiederhake
41460d4c15 cpu-cpuid: Merge addFeature functions
Prepare to deduplicate the list of relevant registers for cpuid and
msr information.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:10:50 +01:00
Tim Wiederhake
5868cfc490 cpu-cpuid: Remove xmltodict usage in parseCPU
'xmltodict' is a Python module that is not installed by default.
Replace it, so the dependencies of cpu-gather.py do not change
when both scripts are merged.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:10:44 +01:00
Tim Wiederhake
dc6e527b2d cpu-cpuid: Remove xmltodict usage in parseMap
'xmltodict' is a Python module that is not installed by default.
Replace it, so the dependencies of cpu-gather.py do not change
when both scripts are merged.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:10:39 +01:00
Tim Wiederhake
b53eb0db35 cpu-cpuid: Use argparse to parse arguments
Using 'argparse' for argument handling simplifies merging this script
with cpu-gather.py in a later patch.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2021-01-07 18:09:38 +01:00
Tim Wiederhake
bed50bcbbb cpu-gather: Add IA32_CORE_CAPABILITY_MSR
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:21:05 +01:00
Tim Wiederhake
942cf935d6 cpu-gather: Prepare gather_msr for reading multiple msr
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:21:00 +01:00
Tim Wiederhake
4b9720e89f cpu-gather: Allow gathering and parsing data in one step.
Make
  ./cpu-gather.py --gather --parse
an alias of
  ./cpu-gather.py [--gather] | ./cpu-gather.py --parse

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:51 +01:00
Tim Wiederhake
731b922a65 cpu-gather: Parse cpuid leaves early
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:49 +01:00
Tim Wiederhake
b0c966de9b cpu-gather: Ignore shutdown messages from qemu
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:48 +01:00
Tim Wiederhake
e18546ed13 cpu-gather: Ignore empty responses from qemu
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:46 +01:00
Tim Wiederhake
bbba6314b7 cpu-parse: Delete old script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:43 +01:00
Tim Wiederhake
0b683f805f cpu-parse: Move call to cpu-cpuid.py to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:41 +01:00
Tim Wiederhake
56d496d926 cpu-parse: Move json output to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:39 +01:00
Tim Wiederhake
58a1ad721f cpu-parse: Move xml output to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:37 +01:00
Tim Wiederhake
43d3703d3a cpu-parse: Move file name generation to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:35 +01:00
Tim Wiederhake
29584c4a48 cpu-parse: Move model name detection to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:32 +01:00
Tim Wiederhake
2933047c02 cpu-gather: Transport data as json
More reliable, easier to parse, easier to edit.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:29 +01:00
Tim Wiederhake
12f01dcb04 cpu-parse: Wrap with python script
This changes the invocation from
  ./cpu-gather.py | ./cpu-parse.sh
to
  ./cpu-gather.py [--gather] | ./cpu-gather.py --parse

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:23 +01:00
Tim Wiederhake
b00a7e87f1 cpu-gather: Separate data input and output
This is a preparatory step to replace the output format with
something more readable.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:21 +01:00
Tim Wiederhake
82878ba049 cpu-gather: Delete old script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:18 +01:00
Tim Wiederhake
d828870d2d cpu-gather: Merge model gathering logic
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:15 +01:00
Tim Wiederhake
b1773967c5 cpu-gather: Move full model extraction to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:11 +01:00
Tim Wiederhake
92a989b9d4 cpu-gather: Move simple model extraction to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:09 +01:00
Tim Wiederhake
afbb7af3ca cpu-gather: Move static model extraction to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:05 +01:00
Tim Wiederhake
090259de55 cpu-gather: Move static model expansion to new script
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-12-18 23:20:03 +01:00