@fugood/llama.node 0.3.1 → 0.3.3

package/src/llama.cpp/ggml/src/{ggml-rpc.cpp → ggml-rpc/ggml-rpc.cpp}

@@ -1,5 +1,5 @@
 #include "ggml-rpc.h"
-#include "ggml.h"
+#include "ggml-impl.h"
 #include "ggml-backend-impl.h"
 
 #include <cinttypes>
@@ -25,7 +25,7 @@
 #  include <netdb.h>
 #  include <unistd.h>
 #endif
-#include <string.h>
+#include <cstring>
 
 #define UNUSED GGML_UNUSED
 
@@ -57,8 +57,9 @@ struct socket_t {
     }
 };
 
-// ggml_tensor is serialized into rpc_tensor
+// all RPC structures must be packed
 #pragma pack(push, 1)
+// ggml_tensor is serialized into rpc_tensor
 struct rpc_tensor {
     uint64_t id;
     uint32_t type;
@@ -76,25 +77,84 @@ struct rpc_tensor {
 
     char padding[4];
 };
-#pragma pack(pop)
 
 static_assert(sizeof(rpc_tensor) % 8 == 0, "rpc_tensor size must be multiple of 8");
 
 // RPC commands
 enum rpc_cmd {
-    ALLOC_BUFFER = 0,
-    GET_ALIGNMENT,
-    GET_MAX_SIZE,
-    BUFFER_GET_BASE,
-    FREE_BUFFER,
-    BUFFER_CLEAR,
-    SET_TENSOR,
-    GET_TENSOR,
-    COPY_TENSOR,
-    GRAPH_COMPUTE,
-    GET_DEVICE_MEMORY,
+    RPC_CMD_ALLOC_BUFFER = 0,
+    RPC_CMD_GET_ALIGNMENT,
+    RPC_CMD_GET_MAX_SIZE,
+    RPC_CMD_BUFFER_GET_BASE,
+    RPC_CMD_FREE_BUFFER,
+    RPC_CMD_BUFFER_CLEAR,
+    RPC_CMD_SET_TENSOR,
+    RPC_CMD_GET_TENSOR,
+    RPC_CMD_COPY_TENSOR,
+    RPC_CMD_GRAPH_COMPUTE,
+    RPC_CMD_GET_DEVICE_MEMORY,
+    RPC_CMD_COUNT,
+};
+
+struct rpc_msg_alloc_buffer_req {
+    uint64_t size;
+};
+
+struct rpc_msg_alloc_buffer_rsp {
+    uint64_t remote_ptr;
+    uint64_t remote_size;
+};
+
+struct rpc_msg_get_alignment_rsp {
+    uint64_t alignment;
+};
+
+struct rpc_msg_get_max_size_rsp {
+    uint64_t max_size;
 };
 
+struct rpc_msg_buffer_get_base_req {
+    uint64_t remote_ptr;
+};
+
+struct rpc_msg_buffer_get_base_rsp {
+    uint64_t base_ptr;
+};
+
+struct rpc_msg_free_buffer_req {
+    uint64_t remote_ptr;
+};
+
+struct rpc_msg_buffer_clear_req {
+    uint64_t remote_ptr;
+    uint8_t value;
+};
+
+struct rpc_msg_get_tensor_req {
+    rpc_tensor tensor;
+    uint64_t offset;
+    uint64_t size;
+};
+
+struct rpc_msg_copy_tensor_req {
+    rpc_tensor src;
+    rpc_tensor dst;
+};
+
+struct rpc_msg_copy_tensor_rsp {
+    uint8_t result;
+};
+
+struct rpc_msg_graph_compute_rsp {
+    uint8_t result;
+};
+
+struct rpc_msg_get_device_memory_rsp {
+    uint64_t free_mem;
+    uint64_t total_mem;
+};
+#pragma pack(pop)
+
 // RPC data structures
 
 static ggml_guid_t ggml_backend_rpc_guid() {
@@ -118,7 +178,6 @@ struct ggml_backend_rpc_buffer_context {
     std::shared_ptr<socket_t> sock;
     std::unordered_map<ggml_backend_buffer_t, void *> base_cache;
     uint64_t remote_ptr;
-    std::string name;
 };
 
 // RPC helper functions
@@ -197,6 +256,10 @@ static std::shared_ptr<socket_t> create_server_socket(const char * host, int por
         fprintf(stderr, "Failed to set SO_REUSEADDR\n");
         return nullptr;
     }
+    if (inet_addr(host) == INADDR_NONE) {
+        fprintf(stderr, "Invalid host address: %s\n", host);
+        return nullptr;
+    }
     struct sockaddr_in serv_addr;
     serv_addr.sin_family = AF_INET;
     serv_addr.sin_addr.s_addr = inet_addr(host);
@@ -235,6 +298,38 @@ static bool recv_data(sockfd_t sockfd, void * data, size_t size) {
     return true;
 }
 
+static bool send_msg(sockfd_t sockfd, const void * msg, size_t msg_size) {
+    if (!send_data(sockfd, &msg_size, sizeof(msg_size))) {
+        return false;
+    }
+    return send_data(sockfd, msg, msg_size);
+}
+
+static bool recv_msg(sockfd_t sockfd, void * msg, size_t msg_size) {
+    uint64_t size;
+    if (!recv_data(sockfd, &size, sizeof(size))) {
+        return false;
+    }
+    if (size != msg_size) {
+        return false;
+    }
+    return recv_data(sockfd, msg, msg_size);
+}
+
+static bool recv_msg(sockfd_t sockfd, std::vector<uint8_t> & input) {
+    uint64_t size;
+    if (!recv_data(sockfd, &size, sizeof(size))) {
+        return false;
+    }
+    try {
+        input.resize(size);
+    } catch (const std::bad_alloc & e) {
+        fprintf(stderr, "Failed to allocate input buffer of size %" PRIu64 "\n", size);
+        return false;
+    }
+    return recv_data(sockfd, input.data(), size);
+}
+
 static bool parse_endpoint(const std::string & endpoint, std::string & host, int & port) {
     size_t pos = endpoint.find(':');
     if (pos == std::string::npos) {
@@ -247,28 +342,27 @@ static bool parse_endpoint(const std::string & endpoint, std::string & host, int
 
 // RPC request : | rpc_cmd (1 byte) | request_size (8 bytes) | request_data (request_size bytes) |
 // RPC response: | response_size (8 bytes) | response_data (response_size bytes) |
-static bool send_rpc_cmd(const std::shared_ptr<socket_t> & sock, enum rpc_cmd cmd, const std::vector<uint8_t> & input, std::vector<uint8_t> & output) {
+static bool send_rpc_cmd(const std::shared_ptr<socket_t> & sock, enum rpc_cmd cmd, const void * input, size_t input_size, void * output, size_t output_size) {
     uint8_t cmd_byte = cmd;
     if (!send_data(sock->fd, &cmd_byte, sizeof(cmd_byte))) {
         return false;
     }
-    uint64_t input_size = input.size();
     if (!send_data(sock->fd, &input_size, sizeof(input_size))) {
         return false;
     }
-    if (!send_data(sock->fd, input.data(), input.size())) {
+    if (!send_data(sock->fd, input, input_size)) {
         return false;
     }
-    uint64_t output_size;
-    if (!recv_data(sock->fd, &output_size, sizeof(output_size))) {
+    // TODO: currently the output_size is always known, do we need support for commands with variable output size?
+    // even if we do, we can skip sending output_size from the server for commands with known output size
+    uint64_t out_size;
+    if (!recv_data(sock->fd, &out_size, sizeof(out_size))) {
         return false;
     }
-    if (output_size == 0) {
-        output.clear();
-        return true;
+    if (out_size != output_size) {
+        return false;
     }
-    output.resize(output_size);
-    if (!recv_data(sock->fd, output.data(), output_size)) {
+    if (!recv_data(sock->fd, output, output_size)) {
         return false;
     }
     return true;
@@ -314,43 +408,26 @@ static std::shared_ptr<socket_t> get_socket(const std::string & endpoint) {
     return sock;
 }
 
-GGML_CALL static const char * ggml_backend_rpc_buffer_get_name(ggml_backend_buffer_t buffer) {
+static void ggml_backend_rpc_buffer_free_buffer(ggml_backend_buffer_t buffer) {
     ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
-    return ctx->name.c_str();
-}
-
-GGML_CALL static void ggml_backend_rpc_buffer_free_buffer(ggml_backend_buffer_t buffer) {
-    ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
-    // input serialization format: | remote_ptr (8 bytes) |
-    std::vector<uint8_t> input(sizeof(uint64_t), 0);
-    uint64_t remote_ptr = ctx->remote_ptr;
-    memcpy(input.data(), &remote_ptr, sizeof(remote_ptr));
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(ctx->sock, FREE_BUFFER, input, output);
+    rpc_msg_free_buffer_req request = {ctx->remote_ptr};
+    bool status = send_rpc_cmd(ctx->sock, RPC_CMD_FREE_BUFFER, &request, sizeof(request), nullptr, 0);
     GGML_ASSERT(status);
-    GGML_ASSERT(output.empty());
     delete ctx;
 }
 
-GGML_CALL static void * ggml_backend_rpc_buffer_get_base(ggml_backend_buffer_t buffer) {
+static void * ggml_backend_rpc_buffer_get_base(ggml_backend_buffer_t buffer) {
     ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
     if (ctx->base_cache.find(buffer) != ctx->base_cache.end()) {
         return ctx->base_cache[buffer];
     }
-    // input serialization format: | remote_ptr (8 bytes) |
-    std::vector<uint8_t> input(sizeof(uint64_t), 0);
-    uint64_t remote_ptr = ctx->remote_ptr;
-    memcpy(input.data(), &remote_ptr, sizeof(remote_ptr));
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(ctx->sock, BUFFER_GET_BASE, input, output);
+    rpc_msg_buffer_get_base_req request = {ctx->remote_ptr};
+    rpc_msg_buffer_get_base_rsp response;
+    bool status = send_rpc_cmd(ctx->sock, RPC_CMD_BUFFER_GET_BASE, &request, sizeof(request), &response, sizeof(response));
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == sizeof(uint64_t));
-    // output serialization format: | base_ptr (8 bytes) |
-    uint64_t base_ptr;
-    memcpy(&base_ptr, output.data(), sizeof(base_ptr));
-    void * base = reinterpret_cast<void *>(base_ptr);
-    ctx->base_cache[buffer] = base;
-    return base;
+    void * base_ptr = reinterpret_cast<void *>(response.base_ptr);
+    ctx->base_cache[buffer] = base_ptr;
+    return base_ptr;
 }
 
 static rpc_tensor serialize_tensor(const ggml_tensor * tensor) {
@@ -383,7 +460,7 @@ static rpc_tensor serialize_tensor(const ggml_tensor * tensor) {
     return result;
 }
 
-GGML_CALL static void ggml_backend_rpc_buffer_init_tensor(ggml_backend_buffer_t buffer, ggml_tensor * tensor) {
+static void ggml_backend_rpc_buffer_init_tensor(ggml_backend_buffer_t buffer, ggml_tensor * tensor) {
     UNUSED(buffer);
     if (ggml_is_quantized(tensor->type)) {
         // TODO: this check is due to MATRIX_ROW_PADDING in CUDA and should be generalized
@@ -391,7 +468,7 @@ GGML_CALL static void ggml_backend_rpc_buffer_init_tensor(ggml_backend_buffer_t
     }
 }
 
-GGML_CALL static void ggml_backend_rpc_buffer_set_tensor(ggml_backend_buffer_t buffer, ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
+static void ggml_backend_rpc_buffer_set_tensor(ggml_backend_buffer_t buffer, ggml_tensor * tensor, const void * data, size_t offset, size_t size) {
     ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
     // input serialization format: | rpc_tensor | offset (8 bytes) | data (size bytes) |
     size_t input_size = sizeof(rpc_tensor) + sizeof(uint64_t) + size;
@@ -400,29 +477,21 @@ GGML_CALL static void ggml_backend_rpc_buffer_set_tensor(ggml_backend_buffer_t b
     memcpy(input.data(), &rpc_tensor, sizeof(rpc_tensor));
     memcpy(input.data() + sizeof(rpc_tensor), &offset, sizeof(offset));
     memcpy(input.data() + sizeof(rpc_tensor) + sizeof(offset), data, size);
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(ctx->sock, SET_TENSOR, input, output);
+    bool status = send_rpc_cmd(ctx->sock, RPC_CMD_SET_TENSOR, input.data(), input.size(), nullptr, 0);
     GGML_ASSERT(status);
 }
 
-GGML_CALL static void ggml_backend_rpc_buffer_get_tensor(ggml_backend_buffer_t buffer, const ggml_tensor * tensor, void * data, size_t offset, size_t size) {
+static void ggml_backend_rpc_buffer_get_tensor(ggml_backend_buffer_t buffer, const ggml_tensor * tensor, void * data, size_t offset, size_t size) {
     ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
-    // input serialization format: | rpc_tensor | offset (8 bytes) | size (8 bytes) |
-    int input_size = sizeof(rpc_tensor) + 2*sizeof(uint64_t);
-    std::vector<uint8_t> input(input_size, 0);
-    rpc_tensor rpc_tensor = serialize_tensor(tensor);
-    memcpy(input.data(), &rpc_tensor, sizeof(rpc_tensor));
-    memcpy(input.data() + sizeof(rpc_tensor), &offset, sizeof(offset));
-    memcpy(input.data() + sizeof(rpc_tensor) + sizeof(offset), &size, sizeof(size));
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(ctx->sock, GET_TENSOR, input, output);
+    rpc_msg_get_tensor_req request;
+    request.tensor = serialize_tensor(tensor);
+    request.offset = offset;
+    request.size = size;
+    bool status = send_rpc_cmd(ctx->sock, RPC_CMD_GET_TENSOR, &request, sizeof(request), data, size);
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == size);
-    // output serialization format: | data (size bytes) |
-    memcpy(data, output.data(), size);
 }
 
-GGML_CALL static bool ggml_backend_rpc_buffer_cpy_tensor(ggml_backend_buffer_t buffer, const ggml_tensor * src, ggml_tensor * dst) {
+static bool ggml_backend_rpc_buffer_cpy_tensor(ggml_backend_buffer_t buffer, const ggml_tensor * src, ggml_tensor * dst) {
     // check if src and dst are on the same server
     ggml_backend_buffer_t src_buffer = src->buffer;
     ggml_backend_rpc_buffer_context * src_ctx = (ggml_backend_rpc_buffer_context *)src_buffer->context;
@@ -432,38 +501,27 @@ GGML_CALL static bool ggml_backend_rpc_buffer_cpy_tensor(ggml_backend_buffer_t b
         return false;
     }
     ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
-    // input serialization format: | rpc_tensor src | rpc_tensor dst |
-    int input_size = 2*sizeof(rpc_tensor);
-    std::vector<uint8_t> input(input_size, 0);
-    rpc_tensor rpc_src = serialize_tensor(src);
-    rpc_tensor rpc_dst = serialize_tensor(dst);
-    memcpy(input.data(), &rpc_src, sizeof(rpc_src));
-    memcpy(input.data() + sizeof(rpc_src), &rpc_dst, sizeof(rpc_dst));
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(ctx->sock, COPY_TENSOR, input, output);
+    rpc_msg_copy_tensor_req request;
+    request.src = serialize_tensor(src);
+    request.dst = serialize_tensor(dst);
+    rpc_msg_copy_tensor_rsp response;
+    bool status = send_rpc_cmd(ctx->sock, RPC_CMD_COPY_TENSOR, &request, sizeof(request), &response, sizeof(response));
     GGML_ASSERT(status);
-    // output serialization format: | result (1 byte) |
-    GGML_ASSERT(output.size() == 1);
-    return output[0];
+    return response.result;
 }
 
-GGML_CALL static void ggml_backend_rpc_buffer_clear(ggml_backend_buffer_t buffer, uint8_t value) {
+static void ggml_backend_rpc_buffer_clear(ggml_backend_buffer_t buffer, uint8_t value) {
     ggml_backend_rpc_buffer_context * ctx = (ggml_backend_rpc_buffer_context *)buffer->context;
-    // serialization format: | bufptr (8 bytes) | value (1 byte) |
-    int input_size = sizeof(uint64_t) + sizeof(uint8_t);
-    std::vector<uint8_t> input(input_size, 0);
-    memcpy(input.data(), &ctx->remote_ptr, sizeof(ctx->remote_ptr));
-    memcpy(input.data() + sizeof(ctx->remote_ptr), &value, sizeof(value));
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(ctx->sock, BUFFER_CLEAR, input, output);
+    rpc_msg_buffer_clear_req request = {ctx->remote_ptr, value};
+    bool status = send_rpc_cmd(ctx->sock, RPC_CMD_BUFFER_CLEAR, &request, sizeof(request), nullptr, 0);
     GGML_ASSERT(status);
 }
 
 static ggml_backend_buffer_i ggml_backend_rpc_buffer_interface = {
-    /* .get_name        = */ ggml_backend_rpc_buffer_get_name,
     /* .free_buffer     = */ ggml_backend_rpc_buffer_free_buffer,
     /* .get_base        = */ ggml_backend_rpc_buffer_get_base,
     /* .init_tensor     = */ ggml_backend_rpc_buffer_init_tensor,
+    /* .memset_tensor   = */ NULL,
     /* .set_tensor      = */ ggml_backend_rpc_buffer_set_tensor,
     /* .get_tensor      = */ ggml_backend_rpc_buffer_get_tensor,
     /* .cpy_tensor      = */ ggml_backend_rpc_buffer_cpy_tensor,
@@ -471,32 +529,23 @@ static ggml_backend_buffer_i ggml_backend_rpc_buffer_interface = {
     /* .reset           = */ NULL,
 };
 
-GGML_CALL static const char * ggml_backend_rpc_buffer_type_name(ggml_backend_buffer_type_t buft) {
+static const char * ggml_backend_rpc_buffer_type_name(ggml_backend_buffer_type_t buft) {
     ggml_backend_rpc_buffer_type_context * buft_ctx = (ggml_backend_rpc_buffer_type_context *)buft->context;
     return buft_ctx->name.c_str();
 }
 
-GGML_CALL static ggml_backend_buffer_t ggml_backend_rpc_buffer_type_alloc_buffer(ggml_backend_buffer_type_t buft, size_t size) {
+static ggml_backend_buffer_t ggml_backend_rpc_buffer_type_alloc_buffer(ggml_backend_buffer_type_t buft, size_t size) {
     ggml_backend_rpc_buffer_type_context * buft_ctx = (ggml_backend_rpc_buffer_type_context *)buft->context;
-    // input serialization format: | size (8 bytes) |
-    int input_size = sizeof(uint64_t);
-    std::vector<uint8_t> input(input_size, 0);
-    memcpy(input.data(), &size, sizeof(size));
-    std::vector<uint8_t> output;
+    rpc_msg_alloc_buffer_req request = {size};
+    rpc_msg_alloc_buffer_rsp response;
     auto sock = get_socket(buft_ctx->endpoint);
-    bool status = send_rpc_cmd(sock, ALLOC_BUFFER, input, output);
+    bool status = send_rpc_cmd(sock, RPC_CMD_ALLOC_BUFFER, &request, sizeof(request), &response, sizeof(response));
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == 2*sizeof(uint64_t));
-    // output serialization format: | remote_ptr (8 bytes) | remote_size (8 bytes) |
-    uint64_t remote_ptr;
-    memcpy(&remote_ptr, output.data(), sizeof(remote_ptr));
-    size_t remote_size;
-    memcpy(&remote_size, output.data() + sizeof(uint64_t), sizeof(remote_size));
-    if (remote_ptr != 0) {
+    if (response.remote_ptr != 0) {
         ggml_backend_buffer_t buffer = ggml_backend_buffer_init(buft,
             ggml_backend_rpc_buffer_interface,
-            new ggml_backend_rpc_buffer_context{sock, {}, remote_ptr, "RPC[" + std::string(buft_ctx->endpoint) + "]"},
-            remote_size);
+            new ggml_backend_rpc_buffer_context{sock, {}, response.remote_ptr},
+            response.remote_size);
         return buffer;
     } else {
         return nullptr;
@@ -504,42 +553,30 @@ GGML_CALL static ggml_backend_buffer_t ggml_backend_rpc_buffer_type_alloc_buffer
 }
 
 static size_t get_alignment(const std::shared_ptr<socket_t> & sock) {
-    // input serialization format: | 0 bytes |
-    std::vector<uint8_t> input;
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(sock, GET_ALIGNMENT, input, output);
+    rpc_msg_get_alignment_rsp response;
+    bool status = send_rpc_cmd(sock, RPC_CMD_GET_ALIGNMENT, nullptr, 0, &response, sizeof(response));
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == sizeof(uint64_t));
-    // output serialization format: | alignment (8 bytes) |
-    uint64_t alignment;
-    memcpy(&alignment, output.data(), sizeof(alignment));
-    return alignment;
+    return response.alignment;
 }
 
-GGML_CALL static size_t ggml_backend_rpc_buffer_type_get_alignment(ggml_backend_buffer_type_t buft) {
+static size_t ggml_backend_rpc_buffer_type_get_alignment(ggml_backend_buffer_type_t buft) {
     ggml_backend_rpc_buffer_type_context * buft_ctx = (ggml_backend_rpc_buffer_type_context *)buft->context;
     return buft_ctx->alignment;
 }
 
 static size_t get_max_size(const std::shared_ptr<socket_t> & sock) {
-    // input serialization format: | 0 bytes |
-    std::vector<uint8_t> input;
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(sock, GET_MAX_SIZE, input, output);
+    rpc_msg_get_max_size_rsp response;
+    bool status = send_rpc_cmd(sock, RPC_CMD_GET_MAX_SIZE, nullptr, 0, &response, sizeof(response));
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == sizeof(uint64_t));
-    // output serialization format: | max_size (8 bytes) |
-    uint64_t max_size;
-    memcpy(&max_size, output.data(), sizeof(max_size));
-    return max_size;
+    return response.max_size;
 }
 
-GGML_CALL static size_t ggml_backend_rpc_get_max_size(ggml_backend_buffer_type_t buft) {
+static size_t ggml_backend_rpc_get_max_size(ggml_backend_buffer_type_t buft) {
     ggml_backend_rpc_buffer_type_context * buft_ctx = (ggml_backend_rpc_buffer_type_context *)buft->context;
     return buft_ctx->max_size;
 }
 
-GGML_CALL static size_t ggml_backend_rpc_buffer_type_get_alloc_size(ggml_backend_buffer_type_t buft, const ggml_tensor * tensor) {
+static size_t ggml_backend_rpc_buffer_type_get_alloc_size(ggml_backend_buffer_type_t buft, const ggml_tensor * tensor) {
     UNUSED(buft);
     return ggml_nbytes(tensor);
 }
@@ -553,24 +590,19 @@ static ggml_backend_buffer_type_i ggml_backend_rpc_buffer_type_interface = {
     /* .is_host          = */ NULL,
 };
 
-GGML_CALL static const char * ggml_backend_rpc_name(ggml_backend_t backend) {
+static const char * ggml_backend_rpc_name(ggml_backend_t backend) {
     ggml_backend_rpc_context * rpc_ctx = (ggml_backend_rpc_context *)backend->context;
 
     return rpc_ctx->name.c_str();
 }
 
-GGML_CALL static void ggml_backend_rpc_free(ggml_backend_t backend) {
+static void ggml_backend_rpc_free(ggml_backend_t backend) {
     ggml_backend_rpc_context * rpc_ctx = (ggml_backend_rpc_context *)backend->context;
     delete rpc_ctx;
     delete backend;
 }
 
-GGML_CALL static ggml_backend_buffer_type_t ggml_backend_rpc_get_default_buffer_type(ggml_backend_t backend) {
-    ggml_backend_rpc_context * ctx = (ggml_backend_rpc_context *)backend->context;
-    return ggml_backend_rpc_buffer_type(ctx->endpoint.c_str());
-}
-
-GGML_CALL static void ggml_backend_rpc_synchronize(ggml_backend_t backend) {
+static void ggml_backend_rpc_synchronize(ggml_backend_t backend) {
     UNUSED(backend);
     // this is no-op because we don't have any async operations
 }
@@ -612,38 +644,20 @@ static void serialize_graph(const ggml_cgraph * cgraph, std::vector<uint8_t> & o
     memcpy(out_tensors, tensors.data(), n_tensors * sizeof(rpc_tensor));
 }
 
-GGML_CALL static enum ggml_status ggml_backend_rpc_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) {
+static enum ggml_status ggml_backend_rpc_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) {
     ggml_backend_rpc_context * rpc_ctx = (ggml_backend_rpc_context *)backend->context;
     std::vector<uint8_t> input;
     serialize_graph(cgraph, input);
-    std::vector<uint8_t> output;
+    rpc_msg_graph_compute_rsp response;
     auto sock = get_socket(rpc_ctx->endpoint);
-    bool status = send_rpc_cmd(sock, GRAPH_COMPUTE, input, output);
+    bool status = send_rpc_cmd(sock, RPC_CMD_GRAPH_COMPUTE, input.data(), input.size(), &response, sizeof(response));
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == 1);
-    return (enum ggml_status)output[0];
-}
-
-GGML_CALL static bool ggml_backend_rpc_supports_op(ggml_backend_t backend, const ggml_tensor * op) {
-    UNUSED(backend);
-    UNUSED(op);
-    //TODO: call the remote backend and cache the results
-    return true;
-}
-
-GGML_CALL static bool ggml_backend_rpc_supports_buft(ggml_backend_t backend, ggml_backend_buffer_type_t buft) {
-    if (buft->iface.get_name != ggml_backend_rpc_buffer_type_name) {
-        return false;
-    }
-    ggml_backend_rpc_buffer_type_context * buft_ctx = (ggml_backend_rpc_buffer_type_context *)buft->context;
-    ggml_backend_rpc_context * rpc_ctx = (ggml_backend_rpc_context *)backend->context;
-    return buft_ctx->endpoint == rpc_ctx->endpoint;
+    return (enum ggml_status)response.result;
 }
 
 static ggml_backend_i ggml_backend_rpc_interface = {
     /* .get_name                = */ ggml_backend_rpc_name,
     /* .free                    = */ ggml_backend_rpc_free,
-    /* .get_default_buffer_type = */ ggml_backend_rpc_get_default_buffer_type,
     /* .set_tensor_async        = */ NULL,
     /* .get_tensor_async        = */ NULL,
     /* .cpy_tensor_async        = */ NULL,
@@ -653,17 +667,11 @@ static ggml_backend_i ggml_backend_rpc_interface = {
     /* .graph_plan_update       = */ NULL,
     /* .graph_plan_compute      = */ NULL,
     /* .graph_compute           = */ ggml_backend_rpc_graph_compute,
-    /* .supports_op             = */ ggml_backend_rpc_supports_op,
-    /* .supports_buft           = */ ggml_backend_rpc_supports_buft,
-    /* .offload_op              = */ NULL,
-    /* .event_new               = */ NULL,
-    /* .event_free              = */ NULL,
     /* .event_record            = */ NULL,
     /* .event_wait              = */ NULL,
-    /* .event_synchronize       = */ NULL,
 };
 
-GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_rpc_buffer_type(const char * endpoint) {
+ggml_backend_buffer_type_t ggml_backend_rpc_buffer_type(const char * endpoint) {
     static std::mutex mutex;
     std::lock_guard<std::mutex> lock(mutex);
     // NOTE: buffer types are allocated and never freed; this is by design
@@ -674,6 +682,7 @@ GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_rpc_buffer_type(const
     }
     auto sock = get_socket(endpoint);
     if (sock == nullptr) {
+        fprintf(stderr, "Failed to connect to %s\n", endpoint);
         return nullptr;
     }
     size_t alignment = get_alignment(sock);
@@ -687,13 +696,14 @@ GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_rpc_buffer_type(const
 
     ggml_backend_buffer_type_t buft = new ggml_backend_buffer_type {
         /* .iface   = */ ggml_backend_rpc_buffer_type_interface,
+        /* .device  = */ ggml_backend_rpc_add_device(endpoint),
         /* .context = */ buft_ctx
     };
     buft_map[endpoint] = buft;
     return buft;
 }
 
-GGML_CALL ggml_backend_t ggml_backend_rpc_init(const char * endpoint) {
+ggml_backend_t ggml_backend_rpc_init(const char * endpoint) {
     ggml_backend_rpc_context * ctx = new ggml_backend_rpc_context {
         /* .endpoint  = */ endpoint,
         /* .name      = */ "RPC[" + std::string(endpoint) + "]",
@@ -702,32 +712,25 @@ GGML_CALL ggml_backend_t ggml_backend_rpc_init(const char * endpoint) {
     ggml_backend_t backend = new ggml_backend {
         /* .guid      = */ ggml_backend_rpc_guid(),
         /* .interface = */ ggml_backend_rpc_interface,
+        /* .device    = */ ggml_backend_rpc_add_device(endpoint),
         /* .context   = */ ctx
     };
     return backend;
 }
 
-GGML_API GGML_CALL bool ggml_backend_is_rpc(ggml_backend_t backend) {
+bool ggml_backend_is_rpc(ggml_backend_t backend) {
     return backend != NULL && ggml_guid_matches(backend->guid, ggml_backend_rpc_guid());
 }
 
 static void get_device_memory(const std::shared_ptr<socket_t> & sock, size_t * free, size_t * total) {
-    // input serialization format: | 0 bytes |
-    std::vector<uint8_t> input;
-    std::vector<uint8_t> output;
-    bool status = send_rpc_cmd(sock, GET_DEVICE_MEMORY, input, output);
+    rpc_msg_get_device_memory_rsp response;
+    bool status = send_rpc_cmd(sock, RPC_CMD_GET_DEVICE_MEMORY, nullptr, 0, &response, sizeof(response));
     GGML_ASSERT(status);
-    GGML_ASSERT(output.size() == 2*sizeof(uint64_t));
-    // output serialization format: | free (8 bytes) | total (8 bytes) |
-    uint64_t free_mem;
-    memcpy(&free_mem, output.data(), sizeof(free_mem));
-    uint64_t total_mem;
-    memcpy(&total_mem, output.data() + sizeof(uint64_t), sizeof(total_mem));
-    *free = free_mem;
-    *total = total_mem;
+    *free = response.free_mem;
+    *total = response.total_mem;
 }
 
-GGML_API GGML_CALL void ggml_backend_rpc_get_device_memory(const char * endpoint, size_t * free, size_t * total) {
+void ggml_backend_rpc_get_device_memory(const char * endpoint, size_t * free, size_t * total) {
     auto sock = get_socket(endpoint);
     if (sock == nullptr) {
         *free = 0;
@@ -744,16 +747,16 @@ public:
     rpc_server(ggml_backend_t backend) : backend(backend) {}
     ~rpc_server();
 
-    bool alloc_buffer(const std::vector<uint8_t> & input, std::vector<uint8_t> & output);
-    void get_alignment(std::vector<uint8_t> & output);
-    void get_max_size(std::vector<uint8_t> & output);
-    bool buffer_get_base(const std::vector<uint8_t> & input, std::vector<uint8_t> & output);
-    bool free_buffer(const std::vector<uint8_t> & input);
-    bool buffer_clear(const std::vector<uint8_t> & input);
+    void alloc_buffer(const rpc_msg_alloc_buffer_req & request, rpc_msg_alloc_buffer_rsp & response);
+    void get_alignment(rpc_msg_get_alignment_rsp & response);
+    void get_max_size(rpc_msg_get_max_size_rsp & response);
+    bool buffer_get_base(const rpc_msg_buffer_get_base_req & request, rpc_msg_buffer_get_base_rsp & response);
+    bool free_buffer(const rpc_msg_free_buffer_req & request);
+    bool buffer_clear(const rpc_msg_buffer_clear_req & request);
     bool set_tensor(const std::vector<uint8_t> & input);
-    bool get_tensor(const std::vector<uint8_t> & input, std::vector<uint8_t> & output);
-    bool copy_tensor(const std::vector<uint8_t> & input, std::vector<uint8_t> & output);
-    bool graph_compute(const std::vector<uint8_t> & input, std::vector<uint8_t> & output);
+    bool get_tensor(const rpc_msg_get_tensor_req & request, std::vector<uint8_t> & response);
+    bool copy_tensor(const rpc_msg_copy_tensor_req & request, rpc_msg_copy_tensor_rsp & response);
+    bool graph_compute(const std::vector<uint8_t> & input, rpc_msg_graph_compute_rsp & response);
 
 private:
     ggml_tensor * deserialize_tensor(struct ggml_context * ctx, const rpc_tensor * tensor);
@@ -767,80 +770,50 @@ private:
     std::unordered_set<ggml_backend_buffer_t> buffers;
 };
 
-bool rpc_server::alloc_buffer(const std::vector<uint8_t> & input, std::vector<uint8_t> & output) {
-    // input serialization format: | size (8 bytes) |
-    if (input.size() != sizeof(uint64_t)) {
-        return false;
-    }
-    uint64_t size;
-    memcpy(&size, input.data(), sizeof(size));
+void rpc_server::alloc_buffer(const rpc_msg_alloc_buffer_req & request, rpc_msg_alloc_buffer_rsp & response) {
     ggml_backend_buffer_type_t buft = ggml_backend_get_default_buffer_type(backend);
-    ggml_backend_buffer_t buffer = ggml_backend_buft_alloc_buffer(buft, size);
-    uint64_t remote_ptr = 0;
-    uint64_t remote_size = 0;
+    ggml_backend_buffer_t buffer = ggml_backend_buft_alloc_buffer(buft, request.size);
+    response.remote_ptr = 0;
+    response.remote_size = 0;
     if (buffer != nullptr) {
-        remote_ptr = reinterpret_cast<uint64_t>(buffer);
-        remote_size = buffer->size;
-        GGML_PRINT_DEBUG("[%s] size: %" PRIu64 " -> remote_ptr: %" PRIx64 ", remote_size: %" PRIu64 "\n", __func__, size, remote_ptr, remote_size);
+        response.remote_ptr = reinterpret_cast<uint64_t>(buffer);
+        response.remote_size = buffer->size;
+        GGML_PRINT_DEBUG("[%s] size: %" PRIu64 " -> remote_ptr: %" PRIx64 ", remote_size: %" PRIu64 "\n", __func__, request.size, response.remote_ptr, response.remote_size);
         buffers.insert(buffer);
     } else {
-        GGML_PRINT_DEBUG("[%s] size: %" PRIu64 " -> failed\n", __func__, size);
+        GGML_PRINT_DEBUG("[%s] size: %" PRIu64 " -> failed\n", __func__, request.size);
     }
-    // output serialization format: | remote_ptr (8 bytes) | remote_size (8 bytes) |
-    output.resize(2*sizeof(uint64_t), 0);
-    memcpy(output.data(), &remote_ptr, sizeof(remote_ptr));
-    memcpy(output.data() + sizeof(uint64_t), &remote_size, sizeof(remote_size));
-    return true;
 }
 
-void rpc_server::get_alignment(std::vector<uint8_t> & output) {
+void rpc_server::get_alignment(rpc_msg_get_alignment_rsp & response) {
     ggml_backend_buffer_type_t buft = ggml_backend_get_default_buffer_type(backend);
     size_t alignment = ggml_backend_buft_get_alignment(buft);
     GGML_PRINT_DEBUG("[%s] alignment: %lu\n", __func__, alignment);
-    // output serialization format: | alignment (8 bytes) |
-    output.resize(sizeof(uint64_t), 0);
-    memcpy(output.data(), &alignment, sizeof(alignment));
+    response.alignment = alignment;
 }
 
-void rpc_server::get_max_size(std::vector<uint8_t> & output) {
+void rpc_server::get_max_size(rpc_msg_get_max_size_rsp & response) {
     ggml_backend_buffer_type_t buft = ggml_backend_get_default_buffer_type(backend);
     size_t max_size = ggml_backend_buft_get_max_size(buft);
     GGML_PRINT_DEBUG("[%s] max_size: %lu\n", __func__, max_size);
-    // output serialization format: | max_size (8 bytes) |
-    output.resize(sizeof(uint64_t), 0);
-    memcpy(output.data(), &max_size, sizeof(max_size));
+    response.max_size = max_size;
 }
 
-bool rpc_server::buffer_get_base(const std::vector<uint8_t> & input, std::vector<uint8_t> & output) {
-    // input serialization format: | remote_ptr (8 bytes) |
-    if (input.size() != sizeof(uint64_t)) {
-        return false;
-    }
-    uint64_t remote_ptr;
-    memcpy(&remote_ptr, input.data(), sizeof(remote_ptr));
-    GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 "\n", __func__, remote_ptr);
-    ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(remote_ptr);
+bool rpc_server::buffer_get_base(const rpc_msg_buffer_get_base_req & request, rpc_msg_buffer_get_base_rsp & response) {
+    GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 "\n", __func__, request.remote_ptr);
+    ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(request.remote_ptr);
     if (buffers.find(buffer) == buffers.end()) {
         GGML_PRINT_DEBUG("[%s] buffer not found\n", __func__);
         return false;
     }
     void * base = ggml_backend_buffer_get_base(buffer);
-    // output serialization format: | base_ptr (8 bytes) |
-    uint64_t base_ptr = reinterpret_cast<uint64_t>(base);
-    output.resize(sizeof(uint64_t), 0);
-    memcpy(output.data(), &base_ptr, sizeof(base_ptr));
+    response.base_ptr = reinterpret_cast<uint64_t>(base);
     return true;
 }
 
-bool rpc_server::free_buffer(const std::vector<uint8_t> & input) {
-    // input serialization format: | remote_ptr (8 bytes) |
-    if (input.size() != sizeof(uint64_t)) {
-        return false;
-    }
-    uint64_t remote_ptr;
-    memcpy(&remote_ptr, input.data(), sizeof(remote_ptr));
-    GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 "\n", __func__, remote_ptr);
-    ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(remote_ptr);
+bool rpc_server::free_buffer(const rpc_msg_free_buffer_req & request) {
+    GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 "\n", __func__, request.remote_ptr);
+    ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(request.remote_ptr);
     if (buffers.find(buffer) == buffers.end()) {
         GGML_PRINT_DEBUG("[%s] buffer not found\n", __func__);
         return false;
@@ -850,22 +823,14 @@ bool rpc_server::free_buffer(const std::vector<uint8_t> & input) {
     return true;
 }
 
-bool rpc_server::buffer_clear(const std::vector<uint8_t> & input) {
-    // input serialization format: | remote_ptr (8 bytes) | value (1 byte) |
-    if (input.size() != sizeof(uint64_t) + sizeof(uint8_t)) {
-        return false;
-    }
-    uint64_t remote_ptr;
-    memcpy(&remote_ptr, input.data(), sizeof(remote_ptr));
-    uint8_t value;
-    memcpy(&value, input.data() + sizeof(uint64_t), sizeof(value));
-    GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 ", value: %u\n", __func__, remote_ptr, value);
-    ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(remote_ptr);
+bool rpc_server::buffer_clear(const rpc_msg_buffer_clear_req & request) {
+    GGML_PRINT_DEBUG("[%s] remote_ptr: %" PRIx64 ", value: %u\n", __func__, request.remote_ptr, request.value);
+    ggml_backend_buffer_t buffer = reinterpret_cast<ggml_backend_buffer_t>(request.remote_ptr);
     if (buffers.find(buffer) == buffers.end()) {
         GGML_PRINT_DEBUG("[%s] buffer not found\n", __func__);
         return false;
     }
-    ggml_backend_buffer_clear(buffer, value);
+    ggml_backend_buffer_clear(buffer, request.value);
     return true;
 }
 
@@ -877,8 +842,18 @@ ggml_tensor * rpc_server::deserialize_tensor(struct ggml_context * ctx, const rp
     }
     result->buffer = reinterpret_cast<ggml_backend_buffer_t>(tensor->buffer);
     if (result->buffer && buffers.find(result->buffer) == buffers.end()) {
-        return nullptr;
+        result->buffer = nullptr;
+    }
+
+    if (result->buffer) {
+        // require that the tensor data does not go beyond the buffer end
+        uint64_t tensor_size = (uint64_t) ggml_nbytes(result);
+        uint64_t buffer_start = (uint64_t) ggml_backend_buffer_get_base(result->buffer);
+        uint64_t buffer_size = (uint64_t) ggml_backend_buffer_get_size(result->buffer);
+        GGML_ASSERT(tensor->data + tensor_size >= tensor->data); // check for overflow
+        GGML_ASSERT(tensor->data >= buffer_start && tensor->data + tensor_size <= buffer_start + buffer_size);
     }
+
     result->op = (ggml_op) tensor->op;
     for (uint32_t i = 0; i < GGML_MAX_OP_PARAMS / sizeof(int32_t); i++) {
         result->op_params[i] = tensor->op_params[i];
@@ -898,7 +873,7 @@ bool rpc_server::set_tensor(const std::vector<uint8_t> & input) {
     const rpc_tensor * in_tensor = (const rpc_tensor *)input.data();
     uint64_t offset;
     memcpy(&offset, input.data() + sizeof(rpc_tensor), sizeof(offset));
-    size_t size = input.size() - sizeof(rpc_tensor) - sizeof(offset);
+    const size_t size = input.size() - sizeof(rpc_tensor) - sizeof(offset);
 
     struct ggml_init_params params {
         /*.mem_size   =*/ ggml_tensor_overhead(),
@@ -913,69 +888,72 @@ bool rpc_server::set_tensor(const std::vector<uint8_t> & input) {
         return false;
     }
     GGML_PRINT_DEBUG("[%s] buffer: %p, data: %p, offset: %" PRIu64 ", size: %zu\n", __func__, (void*)tensor->buffer, tensor->data, offset, size);
+
+    // sanitize tensor->data
+    {
+        const size_t p0 = (size_t) ggml_backend_buffer_get_base(tensor->buffer);
+        const size_t p1 = p0 + ggml_backend_buffer_get_size(tensor->buffer);
+
+        if (in_tensor->data + offset < p0 || in_tensor->data + offset >= p1 || size > (p1 - in_tensor->data - offset)) {
+            GGML_ABORT("[%s] tensor->data out of bounds\n", __func__);
+        }
+    }
+
     const void * data = input.data() + sizeof(rpc_tensor) + sizeof(offset);
     ggml_backend_tensor_set(tensor, data, offset, size);
     ggml_free(ctx);
     return true;
 }
 
-bool rpc_server::get_tensor(const std::vector<uint8_t> & input, std::vector<uint8_t> & output) {
-    // serialization format: | rpc_tensor | offset (8 bytes) | size (8 bytes) |
-    if (input.size() != sizeof(rpc_tensor) + 2*sizeof(uint64_t)) {
-        return false;
-    }
-    const rpc_tensor * in_tensor = (const rpc_tensor *)input.data();
-    uint64_t offset;
-    memcpy(&offset, input.data() + sizeof(rpc_tensor), sizeof(offset));
-    uint64_t size;
-    memcpy(&size, input.data() + sizeof(rpc_tensor) + sizeof(offset), sizeof(size));
-
+bool rpc_server::get_tensor(const rpc_msg_get_tensor_req & request, std::vector<uint8_t> & response) {
     struct ggml_init_params params {
         /*.mem_size   =*/ ggml_tensor_overhead(),
         /*.mem_buffer =*/ NULL,
         /*.no_alloc   =*/ true,
     };
     struct ggml_context * ctx = ggml_init(params);
-    ggml_tensor * tensor = deserialize_tensor(ctx, in_tensor);
+    ggml_tensor * tensor = deserialize_tensor(ctx, &request.tensor);
     if (tensor == nullptr) {
         GGML_PRINT_DEBUG("[%s] error deserializing tensor\n", __func__);
         ggml_free(ctx);
         return false;
     }
-    GGML_PRINT_DEBUG("[%s] buffer: %p, data: %p, offset: %" PRIu64 ", size: %" PRIu64 "\n", __func__, (void*)tensor->buffer, tensor->data, offset, size);
-    // output serialization format: | data (size bytes) |
-    output.resize(size, 0);
-    ggml_backend_tensor_get(tensor, output.data(), offset, size);
+    GGML_PRINT_DEBUG("[%s] buffer: %p, data: %p, offset: %" PRIu64 ", size: %" PRIu64 "\n", __func__, (void*)tensor->buffer, tensor->data, request.offset, request.size);
+
+    // sanitize tensor->data
+    {
+        const size_t p0 = (size_t) ggml_backend_buffer_get_base(tensor->buffer);
+        const size_t p1 = p0 + ggml_backend_buffer_get_size(tensor->buffer);
+
+        if (request.tensor.data + request.offset < p0 ||
+            request.tensor.data + request.offset >= p1 ||
+            request.size > (p1 - request.tensor.data - request.offset)) {
+                GGML_ABORT("[%s] tensor->data out of bounds\n", __func__);
+        }
+    }
+
+    response.resize(request.size, 0);
+    ggml_backend_tensor_get(tensor, response.data(), request.offset, request.size);
     ggml_free(ctx);
     return true;
 }
 
-bool rpc_server::copy_tensor(const std::vector<uint8_t> & input, std::vector<uint8_t> & output) {
-    // serialization format: | rpc_tensor src | rpc_tensor dst |
-    if (input.size() != 2*sizeof(rpc_tensor)) {
-        return false;
-    }
-    const rpc_tensor * rpc_src = (const rpc_tensor *)input.data();
-    const rpc_tensor * rpc_dst = (const rpc_tensor *)(input.data() + sizeof(rpc_src));
-
+bool rpc_server::copy_tensor(const rpc_msg_copy_tensor_req & request, rpc_msg_copy_tensor_rsp & response) {
     struct ggml_init_params params {
         /*.mem_size   =*/ 2*ggml_tensor_overhead(),
         /*.mem_buffer =*/ NULL,
         /*.no_alloc   =*/ true,
     };
     struct ggml_context * ctx = ggml_init(params);
-    ggml_tensor * src = deserialize_tensor(ctx, rpc_src);
-    ggml_tensor * dst = deserialize_tensor(ctx, rpc_dst);
+    ggml_tensor * src = deserialize_tensor(ctx, &request.src);
+    ggml_tensor * dst = deserialize_tensor(ctx, &request.dst);
     if (src == nullptr || dst == nullptr) {
         GGML_PRINT_DEBUG("[%s] error deserializing tensors\n", __func__);
         ggml_free(ctx);
         return false;
     }
     GGML_PRINT_DEBUG("[%s] src->buffer: %p, dst->buffer: %p\n", __func__, (void*)src->buffer, (void*)dst->buffer);
-    bool result = ggml_backend_buffer_copy_tensor(src, dst);
-    // output serialization format: | result (1 byte) |
-    output.resize(1, 0);
-    output[0] = result;
+    response.result = ggml_backend_buffer_copy_tensor(src, dst);
     ggml_free(ctx);
     return true;
 }
@@ -1004,7 +982,7 @@ ggml_tensor * rpc_server::create_node(uint64_t id,
     return result;
 }
 
-bool rpc_server::graph_compute(const std::vector<uint8_t> & input, std::vector<uint8_t> & output) {
+bool rpc_server::graph_compute(const std::vector<uint8_t> & input, rpc_msg_graph_compute_rsp & response) {
     // serialization format:
     // | n_nodes (4 bytes) | nodes (n_nodes * sizeof(uint64_t) | n_tensors (4 bytes) | tensors (n_tensors * sizeof(rpc_tensor)) |
     if (input.size() < sizeof(uint32_t)) {
@@ -1024,7 +1002,7 @@ bool rpc_server::graph_compute(const std::vector<uint8_t> & input, std::vector<u
     const rpc_tensor * tensors = (const rpc_tensor *)(input.data() + sizeof(n_nodes) + n_nodes*sizeof(uint64_t) + sizeof(n_tensors));
     GGML_PRINT_DEBUG("[%s] n_nodes: %u, n_tensors: %u\n", __func__, n_nodes, n_tensors);
 
-    static size_t buf_size = ggml_tensor_overhead()*(n_nodes + n_tensors) + ggml_graph_overhead_custom(n_nodes, false);
+    size_t buf_size = ggml_tensor_overhead()*(n_nodes + n_tensors) + ggml_graph_overhead_custom(n_nodes, false);
     struct ggml_init_params params = {
         /*.mem_size   =*/ buf_size,
         /*.mem_buffer =*/ NULL,
@@ -1044,9 +1022,7 @@ bool rpc_server::graph_compute(const std::vector<uint8_t> & input, std::vector<u
         graph->nodes[i] = create_node(id, ctx, tensor_ptrs, tensor_map);
     }
     ggml_status status = ggml_backend_graph_compute(backend, graph);
-    // output serialization format: | status (1 byte) |
-    output.resize(1, 0);
-    output[0] = status;
+    response.result = status;
     ggml_free(ctx);
     return true;
 }
@@ -1064,84 +1040,162 @@ static void rpc_serve_client(ggml_backend_t backend, sockfd_t sockfd, size_t fre
         if (!recv_data(sockfd, &cmd, 1)) {
             break;
         }
-        std::vector<uint8_t> input;
-        std::vector<uint8_t> output;
-        uint64_t input_size;
-        if (!recv_data(sockfd, &input_size, sizeof(input_size))) {
+        if (cmd >= RPC_CMD_COUNT) {
+            // fail fast if the command is invalid
+            fprintf(stderr, "Unknown command: %d\n", cmd);
             break;
         }
-        input.resize(input_size);
-        if (!recv_data(sockfd, input.data(), input_size)) {
-            break;
-        }
-        bool ok = true;
         switch (cmd) {
-            case ALLOC_BUFFER: {
-                ok = server.alloc_buffer(input, output);
+            case RPC_CMD_ALLOC_BUFFER: {
+                rpc_msg_alloc_buffer_req request;
+                if (!recv_msg(sockfd, &request, sizeof(request))) {
+                    return;
+                }
+                rpc_msg_alloc_buffer_rsp response;
+                server.alloc_buffer(request, response);
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
-            case GET_ALIGNMENT: {
-                server.get_alignment(output);
+            case RPC_CMD_GET_ALIGNMENT: {
+                if (!recv_msg(sockfd, nullptr, 0)) {
+                    return;
+                }
+                rpc_msg_get_alignment_rsp response;
+                server.get_alignment(response);
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
-            case GET_MAX_SIZE: {
-                server.get_max_size(output);
+            case RPC_CMD_GET_MAX_SIZE: {
+                if (!recv_msg(sockfd, nullptr, 0)) {
+                    return;
+                }
+                rpc_msg_get_max_size_rsp response;
+                server.get_max_size(response);
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
-            case BUFFER_GET_BASE: {
-                ok = server.buffer_get_base(input, output);
+            case RPC_CMD_BUFFER_GET_BASE: {
+                rpc_msg_buffer_get_base_req request;
+                if (!recv_msg(sockfd, &request, sizeof(request))) {
+                    return;
+                }
+                rpc_msg_buffer_get_base_rsp response;
+                if (!server.buffer_get_base(request, response)) {
+                    return;
+                }
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
-            case FREE_BUFFER: {
-                ok = server.free_buffer(input);
+            case RPC_CMD_FREE_BUFFER: {
+                rpc_msg_free_buffer_req request;
+                if (!recv_msg(sockfd, &request, sizeof(request))) {
+                    return;
+                }
+                if (!server.free_buffer(request)) {
+                    return;
+                }
+                if (!send_msg(sockfd, nullptr, 0)) {
+                    return;
+                }
                 break;
             }
-            case BUFFER_CLEAR: {
-                ok = server.buffer_clear(input);
+            case RPC_CMD_BUFFER_CLEAR: {
+                rpc_msg_buffer_clear_req request;
+                if (!recv_msg(sockfd, &request, sizeof(request))) {
+                    return;
+                }
+                if (!server.buffer_clear(request)) {
+                    return;
+                }
+                if (!send_msg(sockfd, nullptr, 0)) {
+                    return;
+                }
                 break;
             }
-            case SET_TENSOR: {
-                ok = server.set_tensor(input);
+            case RPC_CMD_SET_TENSOR: {
+                std::vector<uint8_t> input;
+                if (!recv_msg(sockfd, input)) {
+                    return;
+                }
+                if (!server.set_tensor(input)) {
+                    return;
+                }
+                if (!send_msg(sockfd, nullptr, 0)) {
+                    return;
+                }
                 break;
             }
-            case GET_TENSOR: {
-                ok = server.get_tensor(input, output);
+            case RPC_CMD_GET_TENSOR: {
+                rpc_msg_get_tensor_req request;
+                if (!recv_msg(sockfd, &request, sizeof(request))) {
+                    return;
+                }
+                std::vector<uint8_t> response;
+                if (!server.get_tensor(request, response)) {
+                    return;
+                }
+                if (!send_msg(sockfd, response.data(), response.size())) {
+                    return;
+                }
                 break;
             }
-            case COPY_TENSOR: {
-                ok = server.copy_tensor(input, output);
+            case RPC_CMD_COPY_TENSOR: {
+                rpc_msg_copy_tensor_req request;
+                if (!recv_msg(sockfd, &request, sizeof(request))) {
+                    return;
+                }
+                rpc_msg_copy_tensor_rsp response;
+                if (!server.copy_tensor(request, response)) {
+                    return;
+                }
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
-            case GRAPH_COMPUTE: {
-                ok = server.graph_compute(input, output);
+            case RPC_CMD_GRAPH_COMPUTE: {
+                std::vector<uint8_t> input;
+                if (!recv_msg(sockfd, input)) {
+                    return;
+                }
+                rpc_msg_graph_compute_rsp response;
+                if (!server.graph_compute(input, response)) {
+                    return;
+                }
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
-            case GET_DEVICE_MEMORY: {
-                // output serialization format: | free (8 bytes) | total (8 bytes) |
-                output.resize(2*sizeof(uint64_t), 0);
-                memcpy(output.data(), &free_mem, sizeof(free_mem));
-                memcpy(output.data() + sizeof(uint64_t), &total_mem, sizeof(total_mem));
+            case RPC_CMD_GET_DEVICE_MEMORY: {
+                if (!recv_msg(sockfd, nullptr, 0)) {
+                    return;
+                }
+                rpc_msg_get_device_memory_rsp response;
+                response.free_mem = free_mem;
+                response.total_mem = total_mem;
+                if (!send_msg(sockfd, &response, sizeof(response))) {
+                    return;
+                }
                 break;
             }
             default: {
                 fprintf(stderr, "Unknown command: %d\n", cmd);
-                ok = false;
+                return;
             }
         }
-        if (!ok) {
-            break;
-        }
-        uint64_t output_size = output.size();
-        if (!send_data(sockfd, &output_size, sizeof(output_size))) {
-            break;
-        }
-        if (!send_data(sockfd, output.data(), output_size)) {
-            break;
-        }
     }
 }
 
-void start_rpc_server(ggml_backend_t backend, const char * endpoint, size_t free_mem, size_t total_mem) {
+void ggml_backend_rpc_start_server(ggml_backend_t backend, const char * endpoint, size_t free_mem, size_t total_mem) {
     std::string host;
     int port;
     if (!parse_endpoint(endpoint, host, port)) {
@@ -1169,10 +1223,181 @@ void start_rpc_server(ggml_backend_t backend, const char * endpoint, size_t free
             return;
         }
         printf("Accepted client connection, free_mem=%zu, total_mem=%zu\n", free_mem, total_mem);
+        fflush(stdout);
         rpc_serve_client(backend, client_socket->fd, free_mem, total_mem);
         printf("Client connection closed\n");
+        fflush(stdout);
     }
 #ifdef _WIN32
     WSACleanup();
 #endif
 }
+
+// device interface
+
+struct ggml_backend_rpc_device_context {
+    std::string endpoint;
+    std::string name;
+};
+
+static const char * ggml_backend_rpc_device_get_name(ggml_backend_dev_t dev) {
+    ggml_backend_rpc_device_context * ctx = (ggml_backend_rpc_device_context *)dev->context;
+
+    return ctx->name.c_str();
+}
+
+static const char * ggml_backend_rpc_device_get_description(ggml_backend_dev_t dev) {
+    ggml_backend_rpc_device_context * ctx = (ggml_backend_rpc_device_context *)dev->context;
+
+    return ctx->name.c_str();
+}
+
+static void ggml_backend_rpc_device_get_memory(ggml_backend_dev_t dev, size_t * free, size_t * total) {
+    ggml_backend_rpc_device_context * ctx = (ggml_backend_rpc_device_context *)dev->context;
+
+    ggml_backend_rpc_get_device_memory(ctx->endpoint.c_str(), free, total);
+
+    UNUSED(dev);
+}
+
+static enum ggml_backend_dev_type ggml_backend_rpc_device_get_type(ggml_backend_dev_t dev) {
+    // TODO: obtain value from the server
+    return GGML_BACKEND_DEVICE_TYPE_GPU;
+
+    UNUSED(dev);
+}
+
+static void ggml_backend_rpc_device_get_props(ggml_backend_dev_t dev, struct ggml_backend_dev_props * props) {
+    props->name        = ggml_backend_rpc_device_get_name(dev);
+    props->description = ggml_backend_rpc_device_get_description(dev);
+    props->type        = ggml_backend_rpc_device_get_type(dev);
+    ggml_backend_rpc_device_get_memory(dev, &props->memory_free, &props->memory_total);
+    props->caps = {
+        /* .async                 = */ false,
+        /* .host_buffer           = */ false,
+        /* .buffer_from_host_ptr  = */ false,
+        /* .events                = */ false,
+    };
+}
+
+static ggml_backend_t ggml_backend_rpc_device_init(ggml_backend_dev_t dev, const char * params) {
+    ggml_backend_rpc_device_context * ctx = (ggml_backend_rpc_device_context *)dev->context;
+
+    return ggml_backend_rpc_init(ctx->endpoint.c_str());
+
+    UNUSED(params);
+}
+
+static ggml_backend_buffer_type_t ggml_backend_rpc_device_get_buffer_type(ggml_backend_dev_t dev) {
+    ggml_backend_rpc_device_context * ctx = (ggml_backend_rpc_device_context *)dev->context;
+
+    return ggml_backend_rpc_buffer_type(ctx->endpoint.c_str());
+
+    UNUSED(dev);
+}
+
+static bool ggml_backend_rpc_device_supports_op(ggml_backend_dev_t dev, const struct ggml_tensor * op) {
+    UNUSED(dev);
+    UNUSED(op);
+    //TODO: call the remote backend and cache the results
+    return true;
+}
+
+static bool ggml_backend_rpc_device_supports_buft(ggml_backend_dev_t dev, ggml_backend_buffer_type_t buft) {
+    if (!buft || buft->iface.get_name != ggml_backend_rpc_buffer_type_name) {
+        return false;
+    }
+    ggml_backend_rpc_buffer_type_context * buft_ctx = (ggml_backend_rpc_buffer_type_context *)buft->context;
+    ggml_backend_rpc_device_context * dev_ctx = (ggml_backend_rpc_device_context *)dev->context;
+    return buft_ctx->endpoint == dev_ctx->endpoint;
+}
+
+static const struct ggml_backend_device_i ggml_backend_rpc_device_i = {
+    /* .get_name             = */ ggml_backend_rpc_device_get_name,
+    /* .get_description      = */ ggml_backend_rpc_device_get_description,
+    /* .get_memory           = */ ggml_backend_rpc_device_get_memory,
+    /* .get_type             = */ ggml_backend_rpc_device_get_type,
+    /* .get_props            = */ ggml_backend_rpc_device_get_props,
+    /* .init_backend         = */ ggml_backend_rpc_device_init,
+    /* .get_buffer_type      = */ ggml_backend_rpc_device_get_buffer_type,
+    /* .get_host_buffer_type = */ NULL,
+    /* .buffer_from_host_ptr = */ NULL,
+    /* .supports_op          = */ ggml_backend_rpc_device_supports_op,
+    /* .supports_buft        = */ ggml_backend_rpc_device_supports_buft,
+    /* .offload_op           = */ NULL,
+    /* .event_new            = */ NULL,
+    /* .event_free           = */ NULL,
+    /* .event_synchronize    = */ NULL,
+};
+
+// backend reg interface
+
+static const char * ggml_backend_rpc_reg_get_name(ggml_backend_reg_t reg) {
+    return "RPC";
+
+    UNUSED(reg);
+}
+
+static size_t ggml_backend_rpc_reg_get_device_count(ggml_backend_reg_t reg) {
+    return 0;
+
+    UNUSED(reg);
+}
+
+static ggml_backend_dev_t ggml_backend_rpc_reg_get_device(ggml_backend_reg_t reg, size_t index) {
+    GGML_ABORT("The RPC backend does not have enumerated devices - use ggml_backend_add_device instead");
+
+    UNUSED(reg);
+    UNUSED(index);
+}
+
+static void * ggml_backend_rpc_get_proc_address(ggml_backend_reg_t reg, const char * name) {
+    if (std::strcmp(name, "ggml_backend_rpc_add_device") == 0) {
+        return (void *)ggml_backend_rpc_add_device;
+    }
+    return NULL;
+
+    UNUSED(reg);
+}
+
+static const struct ggml_backend_reg_i ggml_backend_rpc_reg_i = {
+    /* .get_name         = */ ggml_backend_rpc_reg_get_name,
+    /* .get_device_count = */ ggml_backend_rpc_reg_get_device_count,
+    /* .get_device       = */ ggml_backend_rpc_reg_get_device,
+    /* .get_proc_address = */ ggml_backend_rpc_get_proc_address,
+};
+
+ggml_backend_reg_t ggml_backend_rpc_reg(void) {
+    static struct ggml_backend_reg ggml_backend_rpc_reg = {
+        /* .iface   = */ ggml_backend_rpc_reg_i,
+        /* .context = */ NULL,
+    };
+
+    return &ggml_backend_rpc_reg;
+}
+
+ggml_backend_dev_t ggml_backend_rpc_add_device(const char * endpoint) {
+    static std::unordered_map<std::string, ggml_backend_dev_t> dev_map;
+
+    static std::mutex mutex;
+    std::lock_guard<std::mutex> lock(mutex);
+
+    if (dev_map.find(endpoint) != dev_map.end()) {
+        return dev_map[endpoint];
+    }
+
+    ggml_backend_rpc_device_context * ctx = new ggml_backend_rpc_device_context {
+        /* .endpoint = */ endpoint,
+        /* .name     = */ "RPC[" + std::string(endpoint) + "]",
+    };
+
+    ggml_backend_dev_t dev = new ggml_backend_device {
+        /* .iface   = */ ggml_backend_rpc_device_i,
+        /* .reg     = */ ggml_backend_rpc_reg(),
+        /* .context = */ ctx,
+    };
+
+    dev_map[endpoint] = dev;
+
+    return dev;
+}