On Sat, Sep 19, 2026 at 13:53:14 +0530, Arun Menon via Devel wrote:
From: Arun Menon <armenon@redhat.com>
SWTPM will support backup and fsync of the state file. [1] and [2] Libvirt needs to support these options while launching a VM. This commit adds these optional attributes to the RNG schema, and updates the logic for both creation and parsing them from the XML.
[1] https://github.com/stefanberger/swtpm/commit/186f72a3a995fc68957ffc0b1438bb6... [2] https://github.com/stefanberger/swtpm/commit/795c1973c913da59925fa6c2d688c5e...
Signed-off-by: Arun Menon <armenon@redhat.com> --- docs/formatdomain.rst | 15 +++++++++++++ src/conf/domain_conf.c | 22 +++++++++++++++++++ src/conf/domain_conf.h | 2 ++ src/conf/schemas/domaincommon.rng | 10 +++++++++ .../qemuxmlconfdata/tpm-emulator-tpm2-enc.xml | 2 +- 5 files changed, 50 insertions(+), 1 deletion(-)
diff --git a/docs/formatdomain.rst b/docs/formatdomain.rst index e4eb2d9ba3..402c3ab221 100644 --- a/docs/formatdomain.rst +++ b/docs/formatdomain.rst @@ -8932,6 +8932,21 @@ Example: usage of external TPM emulator :since:`Since 9.0.0` This attribute only works with the ``emulator`` backend. The accepted values are ``yes`` and ``no``. :since:`Since 7.0.0`
+``backup`` + The optional ``backup`` attribute controls whether swtpm creates a + backup of its TPM state file before updating it. Enabling backup + protects against TPM state file corruption if swtpm process crashes + unexpectedly. Disabling backups avoids file-copying overhead. The + accepted values are ``yes`` and ``no``. :since:`Since 12.8.0`
Since 12.10.0 at this point. As noted before I'm still doubtful that the reasoning of 'file-copying' overhead makes any sense and IMO it should have been by default enabled. Normal users will not enable this if it's not the default until they lose the data. Nevertheless in patch 2/2 it's obvious that swtpm took sub-optimal approach of having just a flag (',backup') rather than an option (',backup=off') so if we were to ever decide that enabling this as default, swtpm will either need to add something like ',no-backup' (look how qemu deprecated these) or would need to happen on libvirt's side.
+ +``fsync`` + The optional ``fsync`` attribute controls whether swtpm explicitly + flushes its state changes to the physical storage using fsync(). + Enabling fsync ensures data remains consistent across host power loss, + but might introduces I/O latency. Disabling fsync removes this latency + at the expense of host-level crash resilience. Acceptable + values are ``yes`` and ``no``. :since:`Since 12.8.0`
12.10.0 (I was slightly more convinced about the argument here, nevertheless it still feels like it should be enabled)
+ ``active_pcr_banks`` The ``active_pcr_banks`` node is used to define which of the PCR banks of a TPM 2.0 to activate. Valid names are for example sha1, sha256, sha384, diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 1327793b0a..283658849a 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -11449,6 +11449,11 @@ virDomainSmartcardDefParseXML(virDomainXMLOption *xmlopt, * </backend> * </tpm> * + * Emulator state backup and fsync is supported with the following: + * <tpm model='tpm-crb'> + * <backend type='emulator' version='2.0' backup='yes' fsync='yes'> + * </tpm> + * */ static virDomainTPMDef * virDomainTPMDefParseXML(virDomainXMLOption *xmlopt, @@ -11562,6 +11567,16 @@ virDomainTPMDefParseXML(virDomainXMLOption *xmlopt, } }
+ if (virXMLPropTristateBool(backends[0], "backup", + VIR_XML_PROP_NONE, + &def->data.emulator.backup) < 0) + goto error; + + if (virXMLPropTristateBool(backends[0], "fsync", + VIR_XML_PROP_NONE, + &def->data.emulator.fsync) < 0) + goto error; + if ((nnodes = virXPathNodeSet("./backend/active_pcr_banks/*", ctxt, &nodes)) < 0) break; if (nnodes > 0) @@ -26743,6 +26758,13 @@ virDomainTPMDefFormat(virBuffer *buf, } if (def->data.emulator.persistent_state) virBufferAddLit(&backendAttrBuf, " persistent_state='yes'"); + + if (def->data.emulator.backup) + virBufferAddLit(&backendAttrBuf, " backup='yes'");
VIR_TRISTATE_BOOL_NO is also non-zero so this will output 'backup='yes' even when the source definition had 'no originally. It needs to use virTristateBoolTypeToString instead of a literal string.
+ + if (def->data.emulator.fsync) + virBufferAddLit(&backendAttrBuf, " fsync='yes'");
..
+ if (def->data.emulator.debug != 0) virBufferAsprintf(&backendAttrBuf, " debug='%u'", def->data.emulator.debug); diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index f7cbe7bf1d..fd022ea66f 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -1546,6 +1546,8 @@ struct _virDomainTPMEmulatorDef { char *name; /* name read from active profile */ virDomainTPMProfileRemoveDisabled removeDisabled; } profile; + virTristateBool backup; + virTristateBool fsync; };
struct _virDomainTPMDef { diff --git a/src/conf/schemas/domaincommon.rng b/src/conf/schemas/domaincommon.rng index 0c0a3597a9..13de7ce528 100644 --- a/src/conf/schemas/domaincommon.rng +++ b/src/conf/schemas/domaincommon.rng @@ -6208,6 +6208,16 @@ <ref name="uint8"/> </attribute> </optional> + <optional> + <attribute name="backup"> + <ref name="virYesNo"/> + </attribute> + </optional> + <optional> + <attribute name="fsync"> + <ref name="virYesNo"/> + </attribute> + </optional> </group> <group> <attribute name="type"> diff --git a/tests/qemuxmlconfdata/tpm-emulator-tpm2-enc.xml b/tests/qemuxmlconfdata/tpm-emulator-tpm2-enc.xml index e6746d5739..515831ab8c 100644 --- a/tests/qemuxmlconfdata/tpm-emulator-tpm2-enc.xml +++ b/tests/qemuxmlconfdata/tpm-emulator-tpm2-enc.xml @@ -28,7 +28,7 @@ <input type='mouse' bus='ps2'/> <input type='keyboard' bus='ps2'/> <tpm model='tpm-tis'> - <backend type='emulator' version='2.0'> + <backend type='emulator' version='2.0' backup='yes' fsync='yes'> <encryption secret='32ee7e76-2178-47a1-ab7b-269e6e348015'/> <source type='dir' path='/some/dir'/> </backend>
Adding another test case where 'no' is shown would make it obvious that 'no' doesn't work.
-- 2.54.0
I'll fix the two bugs and version numbers. No need to send another version. Reviewed-by: Peter Krempa <pkrempa@redhat.com> Nevertheless I'd suggest re-considering the default.