Re: [PATCH v6 2/3] tpm: vtpm_proxy: Implement request_locality function.

From: Stefan Berger
Date: Thu Jun 01 2017 - 20:34:51 EST


On 05/24/2017 08:09 PM, Jarkko Sakkinen wrote:
On Wed, May 24, 2017 at 07:03:27PM -0400, Stefan Berger wrote:
On 05/24/2017 06:21 PM, Jarkko Sakkinen wrote:
On Wed, May 24, 2017 at 05:39:40PM -0400, Stefan Berger wrote:
Implement the request_locality function. To set the locality on the
backend we define vendor-specific TPM 1.2 and TPM 2 ordinals and send
a command to the backend to set the locality for the next commands.

To avoid recursing into requesting the locality, we set the
TPM_TRANSMIT_RAW flag when calling tpm_transmit_cmd. To avoid recursing
into TPM 2 space related commands, we set the space parameter to NULL.

Signed-off-by: Stefan Berger <stefanb@xxxxxxxxxxxxxxxxxx>
---
drivers/char/tpm/tpm-interface.c | 1 +
drivers/char/tpm/tpm_vtpm_proxy.c | 36 ++++++++++++++++++++++++++++++++++++
include/uapi/linux/vtpm_proxy.h | 4 ++++
3 files changed, 41 insertions(+)

diff --git a/drivers/char/tpm/tpm-interface.c b/drivers/char/tpm/tpm-interface.c
index 2eacda2..876d45f 100644
--- a/drivers/char/tpm/tpm-interface.c
+++ b/drivers/char/tpm/tpm-interface.c
@@ -537,6 +537,7 @@ ssize_t tpm_transmit_cmd(struct tpm_chip *chip, struct tpm_space *space,
return 0;
}
+EXPORT_SYMBOL_GPL(tpm_transmit_cmd);
#define TPM_DIGEST_SIZE 20
#define TPM_RET_CODE_IDX 6
diff --git a/drivers/char/tpm/tpm_vtpm_proxy.c b/drivers/char/tpm/tpm_vtpm_proxy.c
index 751059d..66024bf 100644
--- a/drivers/char/tpm/tpm_vtpm_proxy.c
+++ b/drivers/char/tpm/tpm_vtpm_proxy.c
@@ -371,6 +371,41 @@ static bool vtpm_proxy_tpm_req_canceled(struct tpm_chip *chip, u8 status)
return ret;
}
+static int vtpm_proxy_request_locality(struct tpm_chip *chip, int locality)
+{
+ struct tpm_buf buf;
+ int rc;
+ const struct tpm_output_header *header;
+
+ if (chip->flags & TPM_CHIP_FLAG_TPM2)
+ rc = tpm_buf_init(&buf, TPM2_ST_SESSIONS,
+ TPM2_CC_SET_LOCALITY);
I would always go with this branch.

+ else
+ rc = tpm_buf_init(&buf, TPM_TAG_RQU_COMMAND,
+ TPM_ORD_SET_LOCALITY);
+ if (rc)
+ return rc;
+ tpm_buf_append_u8(&buf, locality);
+
+ rc = tpm_transmit_cmd(chip, NULL, buf.data, tpm_buf_length(&buf), 0,
+ TPM_TRANSMIT_UNLOCKED | TPM_TRANSMIT_RAW,
+ "attempting to set locality");
+ if (rc < 0) {
+ locality = rc;
+ goto out;
+ }
+
+ header = (const struct tpm_output_header *)buf.data;
+ rc = be32_to_cpu(header->return_code);
+ if (rc)
+ locality = -1;
+
+out:
+ tpm_buf_destroy(&buf);
+
+ return locality;
+}
+
static const struct tpm_class_ops vtpm_proxy_tpm_ops = {
.flags = TPM_OPS_AUTO_STARTUP,
.recv = vtpm_proxy_tpm_op_recv,
@@ -380,6 +415,7 @@ static const struct tpm_class_ops vtpm_proxy_tpm_ops = {
.req_complete_mask = VTPM_PROXY_REQ_COMPLETE_FLAG,
.req_complete_val = VTPM_PROXY_REQ_COMPLETE_FLAG,
.req_canceled = vtpm_proxy_tpm_req_canceled,
+ .request_locality = vtpm_proxy_request_locality,
};
/*
diff --git a/include/uapi/linux/vtpm_proxy.h b/include/uapi/linux/vtpm_proxy.h
index a69e991..58ac73c 100644
--- a/include/uapi/linux/vtpm_proxy.h
+++ b/include/uapi/linux/vtpm_proxy.h
@@ -46,4 +46,8 @@ struct vtpm_proxy_new_dev {
#define VTPM_PROXY_IOC_NEW_DEV _IOWR(0xa1, 0x00, struct vtpm_proxy_new_dev)
+/* vendor specific commands to set locality */
+#define TPM2_CC_SET_LOCALITY 0x20001000
+#define TPM_ORD_SET_LOCALITY 0x20001000
+
#endif /* _UAPI_LINUX_VTPM_PROXY_H */
--
2.4.3
What practical benefit you get from these two constants? I understand
the value range but not so much the redundant code.
If you want to change this, please go ahead and give it a sinle name.
For the code branches above I think that we should at least send a TPM 1.2
formatted command in case of TPM 1.2 and a TPM 2 formatted one in case of
TPM 2. It seems just 'proper.'

Stefan
Agreed. I will make these changes no need to send new series.

I pushed an update to the test suite, which may help the cause to accept this series.

https://github.com/stefanberger/linux-vtpm-tests

Patch 3/3 should probably not return -EFAULT in case of the user trying to send the SetLocality command via /dev/tpm%d but rather -EPIPE to get the logging of the error suppressed, even though that doesn't seem to be the right error code.


Stefan


Reviewed-by: Jarkko Sakkinen <jarkko.sakkinen@xxxxxxxxxxxxxxx>

/Jarkko