Abstract: An apparatus to facilitate enabling stateless accelerator designs shared across mutually-distrustful tenants is disclosed. The apparatus includes a fully-homomorphic encryption (FHE)-capable circuitry to establish a secure session with a trusted environment executing on a host device communicably coupled to the apparatus; generate, as part of establishing the secure session, per-tenant FHE keys for each tenant utilizing the FHE-capable circuitry, the per-tenant FHE keys utilized to encrypt tenant data provided to an FHE-capable compute kernel of the FHE-capable circuitry; process tenant data that is in an FHE-encrypted format encrypted with a per-tenant FHE key of the per-tenant FHE keys; and store the tenant data that is in the FHE-encrypted format encrypted with the per-tenant FHE key of the per-tenant FHE keys.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes speculation vulnerability detection hardware and execution hardware. The speculation vulnerability detection hardware is to detect vulnerability to a speculative execution attack and, in connection with a detection of vulnerability to a speculative execution attack, to provide an indication that data from a first operation is tainted. The execution hardware is to perform a second operation using the data if the second operation is to be performed non-speculatively and to prevent performance of the second operation if the second operation is to be performed speculatively and the data is tainted.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes decode circuitry and execution circuitry coupled to the decode circuitry. The decode circuitry is to decode a register hardening instruction to mitigate vulnerability to a speculative execution attack. The execution circuitry is to be hardened in response to the register hardening instruction.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes a decode circuitry and store circuitry coupled to the decode circuitry. The decode circuitry is to decode a store hardening instruction to mitigate vulnerability to a speculative execution attack. The store circuitry is to be hardened in response to the store hardening instruction.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes decode circuitry and branch circuitry coupled to the decode circuitry. The decode circuitry is to decode a branch hardening instruction to mitigate vulnerability to a speculative execution attack. The branch circuitry is to be hardened in response to the branch hardening instruction.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes a hybrid key generator and memory protection hardware. The hybrid key generator is to generate a hybrid key based on a public key and multiple process identifiers. Each of the process identifiers corresponds to one or more memory spaces in a memory. The memory protection hardware is to use the first hybrid key to protect to the memory spaces.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes decode circuitry and execution circuitry coupled to the decode circuitry. The decode circuitry is to decode a single instruction to mitigate vulnerability to a speculative execution attack. The execution circuitry is to be hardened in response to the single instruction.

Abstract: Embodiments for dynamically mitigating speculation vulnerabilities are disclosed. In an embodiment, an apparatus includes decode circuitry and load circuitry coupled to the decode circuitry. The decode circuitry is to decode a load hardening instruction to mitigate vulnerability to a speculative execution attack. The load circuitry is to be hardened in response to the load hardening instruction.

Abstract: Techniques and mechanisms for a victim cache to operate in conjunction with a skewed cache to help mitigate the risk of a side-channel attack. In an embodiment, a first line is evicted from a skewed cache, and moved to a victim cache, based on a message indicating that a second line is to be stored to the skewed cache. Subsequently, a request to access the first line results in a search of both the victim cache and sets of the skewed cache which have been mapped to an address corresponding to the first line. Based on the search, the first line is evicted from the victim cache, and reinserted in the skewed cache. In another embodiment, reinsertion of the first line in the skewed cache includes the first line and a third line being swapped between the skewed cache and the victim cache.

Abstract: An apparatus to facilitate enabling stateless accelerator designs shared across mutually-distrustful tenants is disclosed. The apparatus includes a fully-homomorphic encryption (FHE)-capable compute kernel. The FHE-capable compute kernel is to establish a secure session with a trusted environment executing on a host device communicably coupled to the apparatus; generate, as part of establishing the secure session, per-tenant FHE keys for each tenant utilizing the FHE-capable compute kernel, the per-tenant FHE keys utilized to encrypt tenant data provided to the FHE-capable compute kernel; process tenant data that is in an FHE-encrypted format encrypted with a per-tenant FHE key of the per-tenant FHE keys; and store the tenant data that is in the FHE-encrypted format encrypted with the per-tenant FHE key of the per-tenant FHE keys.

Abstract: In one embodiment, a processor includes: a decode circuit to decode a load instruction that is to load an operand to a destination register, the decode circuit to generate at least one fencing micro-operation (?op) associated with the destination register; and a scheduler circuit coupled to the decode circuit. The scheduler circuit is to prevent speculative execution of one or more instructions that consume the operand in response to the at least one fencing ?op. Other embodiments are described and claimed.

Abstract: An apparatus and method for tracking speculative execution flow and detecting potential vulnerabilities.

Abstract: An apparatus and method for non-speculative resource deallocation.

Abstract: An apparatus to facilitate enabling stateless accelerator designs shared across mutually-distrustful tenants is disclosed. The apparatus includes a fully-homomorphic encryption (FHE)-capable compute kernel. The FHE-capable compute kernel is to establish a secure session with a trusted environment executing on a host device communicably coupled to the apparatus; generate, as part of establishing the secure session, per-tenant FHE keys for each tenant utilizing the FHE-capable compute kernel, the per-tenant FHE keys utilized to encrypt tenant data provided to the FHE-capable compute kernel; process tenant data that is in an FHE-encrypted format encrypted with a per-tenant FHE key of the per-tenant FHE keys; and store the tenant data that is in the FHE-encrypted format encrypted with the per-tenant FHE key of the per-tenant FHE keys.

Abstract: A processor includes a processing core to identify a code comprising a plurality of instructions to be executed in the architecturally-protected environment, determine that a first physical memory page stored in the architecturally-protected memory matches a first virtual memory page referenced by a first instruction of the plurality of instructions, generate a first address mapping between a first address of the first virtual memory page and a second address of the first physical memory page, store, in the cache memory, the address translation data structure comprising the first address mapping, and execute the code by retrieving the first address mapping in the address translation data structures to be executed in the architecturally-protected environment, determine that a first physical memory page stored in the architecturally-protected memory matches a first virtual memory page referenced by a first instruction of the plurality of instructions, generate a first address mapping between a first address of

Abstract: Detailed herein are systems, apparatuses, and methods for a computer architecture with instruction set support to mitigate against page fault- and/or cache-based side-channel attacks. In an embodiment, an apparatus includes a decoder to decode a first instruction, the first instruction having a first field for a first opcode that indicates that execution circuitry is to set a first flag in a first register that indicates a mode of operation that redirects program flow to an exception handler upon the occurrence of an event. The apparatus further includes execution circuitry to execute the decoded first instruction to set the first flag in the first register that indicates the mode of operation and to store an address of an exception handler in a second register.

Abstract: Embodiments of methods and apparatuses for hardware load hardening are disclosed. In an embodiment, a processor includes safe logic, data forwarding hardware, and data fetching hardware. The safe logic is to determine whether a load is safe. The data forwarding hardware is to, in response to a determination that the load is safe, forward data requested by the load. The data fetching logic is to fetch the data requested by the load, regardless of the determination that the load is safe.

Abstract: Detailed herein are systems, apparatuses, and methods for a computer architecture with instruction set support to mitigate against page fault and/or cache-based side-channel attacks. In an embodiment, a processor includes a decoder to decode an instruction into a decoded instruction, the instruction comprising a first field that indicates an instruction pointer to a user-level event handler; and an execution unit to execute the decoded instruction to, after a swap of an instruction pointer that indicates where an event occurred from a current instruction pointer register into a user-level event handler pointer register, push the instruction pointer that indicates where the event occurred onto call stack storage, and change a current instruction pointer in the current instruction pointer register to the instruction pointer to the user-level event handler.

Abstract: Embodiments of methods and apparatuses for restricted speculative execution are disclosed. In an embodiment, a processor includes configuration storage, an execution circuit, and a controller. The configuration storage is to store an indicator to enable a restricted speculative execution mode of operation of the processor, wherein the processor is to restrict speculative execution when operating in restricted speculative execution mode. The execution circuit is to perform speculative execution. The controller to restrict speculative execution by the execution circuit when the restricted speculative execution mode is enabled.

Abstract: Detailed herein are systems, apparatuses, and methods for a computer architecture with instruction set support to mitigate against page fault- and/or cache-based side-channel attacks. In an embodiment, an apparatus includes a decoder to decode a first instruction, the first instruction having a first field for a first opcode that indicates that execution circuitry is to set a first flag in a first register that indicates a mode of operation that redirects program flow to an exception handler upon the occurrence of an event. The apparatus further includes execution circuitry to execute the decoded first instruction to set the first flag in the first register that indicates the mode of operation and to store an address of an exception handler in a second register.