Hello, I would like to know the kind of information that a Trusted Application know about the client applications that interact with them? Is there a way for a TA to distinguish between two client applications?
If the TA has no knowledge about CAs, how about the OPTEE kernel? Does the kernel know any information that can differentiate between client applications??
Thank you, Francis
Hi Francis,
On Fri, Apr 13, 2018 at 10:03 AM, Francis Akowuah francisakowuah@gmail.com wrote:
Hello, I would like to know the kind of information that a Trusted Application know about the client applications that interact with them? Is there a way for a TA to distinguish between two client applications?
In the current configuration, no. However, there is a TEE_Identity type describing the CA which the TA can access from the property "gpd.client.identity". The problem is just that it's not entirely defined what should go into TEE_Identity more than it must not have TEE_LOGIN_TRUSTED_APP in .login if the client is in normal world. It's basically up to the Linux kernel driver to fill in some meaningful information in the .uuid field.
If the TA has no knowledge about CAs, how about the OPTEE kernel? Does the kernel know any information that can differentiate between client applications??
The kernel know which process it is etc.
Regards, Jens
-
Francis Akowuah -
Jens Wiklander