From: Stefan Berger <stefanb@linux.ibm.com>
To: David Gibson <david@gibson.dropbear.id.au>,
Stefan Berger <stefanb@linux.vnet.ibm.com>
Cc: marcandre.lureau@redhat.com, qemu-ppc@nongnu.org, qemu-devel@nongnu.org
Subject: Re: [PATCH v7 4/6] tpm_spapr: Support suspend and resume
Date: Fri, 3 Jan 2020 10:45:37 -0500 [thread overview]
Message-ID: <10087fcb-f2c2-8f64-95af-412543fadb9a@linux.ibm.com> (raw)
In-Reply-To: <20200103001900.GM2098@umbus>
On 1/2/20 7:19 PM, David Gibson wrote:
> On Thu, Dec 19, 2019 at 09:06:03AM -0500, Stefan Berger wrote:
>> Extend the tpm_spapr frontend with VM suspend and resume support.
>>
>> Signed-off-by: Stefan Berger <stefanb@linux.vnet.ibm.com>
>> ---
>> hw/tpm/tpm_spapr.c | 67 ++++++++++++++++++++++++++++++++++++++++++++-
>> hw/tpm/trace-events | 2 ++
>> 2 files changed, 68 insertions(+), 1 deletion(-)
>>
>> diff --git a/hw/tpm/tpm_spapr.c b/hw/tpm/tpm_spapr.c
>> index ab184fbb82..cf5c7851e7 100644
>> --- a/hw/tpm/tpm_spapr.c
>> +++ b/hw/tpm/tpm_spapr.c
>> @@ -76,6 +76,9 @@ typedef struct {
>>
>> unsigned char buffer[TPM_SPAPR_BUFFER_MAX];
>>
>> + uint32_t numbytes; /* number of bytes in suspend_buffer */
>> + unsigned char *suspend_buffer;
>> +
>> TPMBackendCmd cmd;
>>
>> TPMBackend *be_driver;
>> @@ -240,6 +243,13 @@ static void tpm_spapr_request_completed(TPMIf *ti, int ret)
>>
>> /* a max. of be_buffer_size bytes can be transported */
>> len = MIN(tpm_cmd_get_size(s->buffer), s->be_buffer_size);
>> +
>> + if (runstate_check(RUN_STATE_FINISH_MIGRATE)) {
>> + /* defer delivery of response until .post_load */
>> + s->numbytes = len;
>> + return;
>> + }
> I'm not understanding the basics of what's going on here. IIUC, the
> backend TPM can complete a request after we've entered migration
The tpm_backend_finish_sync() in the .pre_save function will delay the
writing of the device state until all outstanding responses have been
captured. In case a response was outstanding, the above s->numbytes will
then be set and looked at after return from tpm_backend_finish_sync() --
see below.
> downtime. But if that's the case, I can't see any guarantee that we
> won't have already transmitted the TPM device state, so updating it
> here might never reach the destination. In that case we'd still lose
> the request, so I'm not sure what we're accomplishing here.
>
>> rc = spapr_vio_dma_write(&s->vdev, be32_to_cpu(crq->data),
>> s->buffer, len);
>>
>> @@ -288,11 +298,13 @@ static void tpm_spapr_reset(SpaprVioDevice *dev)
>> SpaprTpmState *s = VIO_SPAPR_VTPM(dev);
>>
>> s->state = SPAPR_VTPM_STATE_NONE;
>> + s->numbytes = 0;
>>
>> s->be_tpm_version = tpm_backend_get_tpm_version(s->be_driver);
>>
>> s->be_buffer_size = MIN(tpm_backend_get_buffer_size(s->be_driver),
>> TPM_SPAPR_BUFFER_MAX);
>> + s->suspend_buffer = g_realloc(s->suspend_buffer, s->be_buffer_size);
>>
>> tpm_backend_reset(s->be_driver);
>> tpm_spapr_do_startup_tpm(s, s->be_buffer_size);
>> @@ -309,9 +321,62 @@ static enum TPMVersion tpm_spapr_get_version(TPMIf *ti)
>> return tpm_backend_get_tpm_version(s->be_driver);
>> }
>>
>> +/* persistent state handling */
>> +
>> +static int tpm_spapr_pre_save(void *opaque)
>> +{
>> + SpaprTpmState *s = opaque;
>> +
>> + tpm_backend_finish_sync(s->be_driver);
>> +
>> + if (s->numbytes) {
>> + memcpy(s->suspend_buffer, s->buffer, s->numbytes);
>> + }
>> +
>> + trace_tpm_spapr_pre_save(s->numbytes);
>> + /*
>> + * we cannot deliver the results to the VM since DMA would touch VM memory
>> + */
>> +
>> + return 0;
>> +}
>> +
>> +static int tpm_spapr_post_load(void *opaque, int version_id)
>> +{
>> + SpaprTpmState *s = opaque;
>> +
>> + if (s->numbytes) {
>> + trace_tpm_spapr_post_load();
>> +
>> + memcpy(s->buffer, s->suspend_buffer,
>> + MIN(s->numbytes, s->be_buffer_size));
>> + /* deliver the results to the VM via DMA */
>> + tpm_spapr_request_completed(TPM_IF(s), 0);
>> + s->numbytes = 0;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> static const VMStateDescription vmstate_spapr_vtpm = {
>> .name = "tpm-spapr",
>> - .unmigratable = 1,
>> + .version_id = 1,
>> + .minimum_version_id = 0,
>> + .minimum_version_id_old = 0,
>> + .pre_save = tpm_spapr_pre_save,
>> + .post_load = tpm_spapr_post_load,
>> + .fields = (VMStateField[]) {
>> + VMSTATE_SPAPR_VIO(vdev, SpaprTpmState),
>> +
>> + VMSTATE_UINT8(state, SpaprTpmState),
>> + VMSTATE_UINT32(numbytes, SpaprTpmState),
>> + VMSTATE_VBUFFER_ALLOC_UINT32(suspend_buffer,
>> + SpaprTpmState, 0, NULL,
>> + numbytes),
>> + /* remember DMA address */
>> + VMSTATE_UINT32(crq.data, SpaprTpmState),
>> + VMSTATE_END_OF_LIST(),
>> + }
>> };
>>
>> static Property tpm_spapr_properties[] = {
>> diff --git a/hw/tpm/trace-events b/hw/tpm/trace-events
>> index 9143a8eaa3..5592cec7de 100644
>> --- a/hw/tpm/trace-events
>> +++ b/hw/tpm/trace-events
>> @@ -67,3 +67,5 @@ tpm_spapr_do_crq_get_version(uint32_t version) "response: version %u"
>> tpm_spapr_do_crq_prepare_to_suspend(void) "response: preparing to suspend"
>> tpm_spapr_do_crq_unknown_msg_type(uint8_t type) "Unknown message type 0x%02x"
>> tpm_spapr_do_crq_unknown_crq(uint8_t raw1, uint8_t raw2) "unknown CRQ 0x%02x 0x%02x ..."
>> +tpm_spapr_pre_save(uint32_t v) "Number of TPM response bytes to deliver after resume: %u"
>> +tpm_spapr_post_load(void) "Delivering TPM response after resume"
next prev parent reply other threads:[~2020-01-03 15:47 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-19 14:05 [PATCH v7 0/6] Add vTPM emulator support for ppc64 platform Stefan Berger
2019-12-19 14:06 ` [PATCH v7 1/6] tpm: Move tpm_tis_show_buffer to tpm_util.c Stefan Berger
2019-12-19 14:29 ` Philippe Mathieu-Daudé
2019-12-20 8:23 ` David Gibson
2019-12-20 12:42 ` Stefan Berger
2019-12-19 14:06 ` [PATCH v7 2/6] spapr: Implement get_dt_compatible() callback Stefan Berger
2019-12-19 14:06 ` [PATCH v7 3/6] tpm_spapr: Support TPM for ppc64 using CRQ based interface Stefan Berger
2019-12-19 14:06 ` [PATCH v7 4/6] tpm_spapr: Support suspend and resume Stefan Berger
2020-01-03 0:19 ` David Gibson
2020-01-03 15:45 ` Stefan Berger [this message]
2019-12-19 14:06 ` [PATCH v7 5/6] hw/ppc/Kconfig: Enable TPM_SPAPR as part of PSERIES config Stefan Berger
2019-12-19 14:06 ` [PATCH v7 6/6] docs: tpm: Add example command line for ppc64 and tpm-spapr Stefan Berger
2020-01-03 0:20 ` David Gibson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=10087fcb-f2c2-8f64-95af-412543fadb9a@linux.ibm.com \
--to=stefanb@linux.ibm.com \
--cc=david@gibson.dropbear.id.au \
--cc=marcandre.lureau@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
--cc=stefanb@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).