tpm: Wrap the buffer from the caller to tpm_buf in tpm_send()

tpm_send() does not give anymore the result back to the caller. This
would require another memcpy(), which kind of tells that the whole
approach is somewhat broken. Instead, as Mimi suggested, this commit
just wraps the data to the tpm_buf, and thus the result will not go to
the garbage.

Obviously this assumes from the caller that it passes large enough
buffer, which makes the whole API somewhat broken because it could be
different size than @buflen but since trusted keys is the only module
using this API right now I think that this fix is sufficient for the
moment.

In the near future the plan is to replace the parameters with a tpm_buf
created by the caller.

Reported-by: Mimi Zohar <zohar@linux.ibm.com>
Suggested-by: Mimi Zohar <zohar@linux.ibm.com>
Cc: stable@vger.kernel.org
Fixes: 412eb585587a ("use tpm_buf in tpm_transmit_cmd() as the IO parameter")
Signed-off-by: Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>
Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>

1 files changed, 2 insertions, 7 deletions

diff --git a/drivers/char/tpm/tpm-interface.c b/drivers/char/tpm/tpm-interface.c
index d9ace5480665..d7a3888ad80f 100644
--- a/drivers/char/tpm/tpm-interface.c
+++ b/drivers/char/tpm/tpm-interface.c
@@ -358,14 +358,9 @@ int tpm_send(struct tpm_chip *chip, void *cmd, size_t buflen)
 	if (!chip)
 		return -ENODEV;
 
-	rc = tpm_buf_init(&buf, 0, 0);
-	if (rc)
-		goto out;
-
-	memcpy(buf.data, cmd, buflen);
+	buf.data = cmd;
 	rc = tpm_transmit_cmd(chip, &buf, 0, "attempting to a send a command");
-	tpm_buf_destroy(&buf);
-out:
+
 	tpm_put_ops(chip);
 	return rc;
 }