The resource manager is a special virtual machine which is always
running on a Gunyah system. It provides APIs for creating and destroying
VMs, secure memory management, sharing/lending of memory between VMs,
and setup of inter-VM communication. Calls to the resource manager are
made via message queues.
This patch implements the basic probing and RPC mechanism to make those
API calls. Request/response calls can be made with gh_rm_call.
Drivers can also register to notifications pushed by RM via
gh_rm_register_notifier
Specific API calls that resource manager supports will be implemented in
subsequent patches.
Signed-off-by: Elliot Berman <[email protected]>
---
drivers/virt/gunyah/Makefile | 4 +-
drivers/virt/gunyah/rsc_mgr.c | 724 ++++++++++++++++++++++++++++++++++++++++++
drivers/virt/gunyah/rsc_mgr.h | 28 ++
3 files changed, 755 insertions(+), 1 deletion(-)
diff --git a/drivers/virt/gunyah/Makefile b/drivers/virt/gunyah/Makefile
index 34f32110faf9..c2308389f551 100644
--- a/drivers/virt/gunyah/Makefile
+++ b/drivers/virt/gunyah/Makefile
@@ -1,3 +1,5 @@
# SPDX-License-Identifier: GPL-2.0
-obj-$(CONFIG_GUNYAH) += gunyah.o
+gunyah_rsc_mgr-y += rsc_mgr.o
+
+obj-$(CONFIG_GUNYAH) += gunyah.o gunyah_rsc_mgr.o
diff --git a/drivers/virt/gunyah/rsc_mgr.c b/drivers/virt/gunyah/rsc_mgr.c
new file mode 100644
index 000000000000..a3578a0c10b4
--- /dev/null
+++ b/drivers/virt/gunyah/rsc_mgr.c
@@ -0,0 +1,724 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (c) 2022-2024 Qualcomm Innovation Center, Inc. All rights reserved.
+ */
+
+#include <linux/completion.h>
+#include <linux/gunyah.h>
+#include <linux/module.h>
+#include <linux/mutex.h>
+#include <linux/notifier.h>
+#include <linux/of.h>
+#include <linux/platform_device.h>
+
+#include "rsc_mgr.h"
+
+/* clang-format off */
+#define RM_RPC_API_VERSION_MASK GENMASK(3, 0)
+#define RM_RPC_HEADER_WORDS_MASK GENMASK(7, 4)
+#define RM_RPC_API_VERSION FIELD_PREP(RM_RPC_API_VERSION_MASK, 1)
+#define RM_RPC_HEADER_WORDS FIELD_PREP(RM_RPC_HEADER_WORDS_MASK, \
+ (sizeof(struct gunyah_rm_rpc_hdr) / sizeof(u32)))
+#define RM_RPC_API (RM_RPC_API_VERSION | RM_RPC_HEADER_WORDS)
+
+#define RM_RPC_TYPE_CONTINUATION 0x0
+#define RM_RPC_TYPE_REQUEST 0x1
+#define RM_RPC_TYPE_REPLY 0x2
+#define RM_RPC_TYPE_NOTIF 0x3
+#define RM_RPC_TYPE_MASK GENMASK(1, 0)
+
+#define GUNYAH_RM_MAX_NUM_FRAGMENTS 62
+#define RM_RPC_FRAGMENTS_MASK GENMASK(7, 2)
+/* clang-format on */
+
+struct gunyah_rm_rpc_hdr {
+ u8 api;
+ u8 type;
+ __le16 seq;
+ __le32 msg_id;
+} __packed;
+
+struct gunyah_rm_rpc_reply_hdr {
+ struct gunyah_rm_rpc_hdr hdr;
+ __le32 err_code; /* GUNYAH_RM_ERROR_* */
+} __packed;
+
+#define GUNYAH_RM_MSGQ_MSG_SIZE 240
+#define GUNYAH_RM_PAYLOAD_SIZE \
+ (GUNYAH_RM_MSGQ_MSG_SIZE - sizeof(struct gunyah_rm_rpc_hdr))
+
+/* RM Error codes */
+enum gunyah_rm_error {
+ /* clang-format off */
+ GUNYAH_RM_ERROR_OK = 0x0,
+ GUNYAH_RM_ERROR_UNIMPLEMENTED = 0xFFFFFFFF,
+ GUNYAH_RM_ERROR_NOMEM = 0x1,
+ GUNYAH_RM_ERROR_NORESOURCE = 0x2,
+ GUNYAH_RM_ERROR_DENIED = 0x3,
+ GUNYAH_RM_ERROR_INVALID = 0x4,
+ GUNYAH_RM_ERROR_BUSY = 0x5,
+ GUNYAH_RM_ERROR_ARGUMENT_INVALID = 0x6,
+ GUNYAH_RM_ERROR_HANDLE_INVALID = 0x7,
+ GUNYAH_RM_ERROR_VALIDATE_FAILED = 0x8,
+ GUNYAH_RM_ERROR_MAP_FAILED = 0x9,
+ GUNYAH_RM_ERROR_MEM_INVALID = 0xA,
+ GUNYAH_RM_ERROR_MEM_INUSE = 0xB,
+ GUNYAH_RM_ERROR_MEM_RELEASED = 0xC,
+ GUNYAH_RM_ERROR_VMID_INVALID = 0xD,
+ GUNYAH_RM_ERROR_LOOKUP_FAILED = 0xE,
+ GUNYAH_RM_ERROR_IRQ_INVALID = 0xF,
+ GUNYAH_RM_ERROR_IRQ_INUSE = 0x10,
+ GUNYAH_RM_ERROR_IRQ_RELEASED = 0x11,
+ /* clang-format on */
+};
+
+/**
+ * struct gunyah_rm_message - Represents a complete message from resource manager
+ * @payload: Combined payload of all the fragments (msg headers stripped off).
+ * @size: Size of the payload received so far.
+ * @msg_id: Message ID from the header.
+ * @type: RM_RPC_TYPE_REPLY or RM_RPC_TYPE_NOTIF.
+ * @num_fragments: total number of fragments expected to be received.
+ * @fragments_received: fragments received so far.
+ * @reply: Fields used for request/reply sequences
+ */
+struct gunyah_rm_message {
+ void *payload;
+ size_t size;
+ u32 msg_id;
+ u8 type;
+
+ u8 num_fragments;
+ u8 fragments_received;
+
+ /**
+ * @ret: Linux return code, there was an error processing message
+ * @seq: Sequence ID for the main message.
+ * @rm_error: For request/reply sequences with standard replies
+ * @seq_done: Signals caller that the RM reply has been received
+ */
+ struct {
+ int ret;
+ u16 seq;
+ enum gunyah_rm_error rm_error;
+ struct completion seq_done;
+ } reply;
+};
+
+/**
+ * struct gunyah_rm - private data for communicating w/Gunyah resource manager
+ * @dev: pointer to RM platform device
+ * @tx_ghrsc: message queue resource to TX to RM
+ * @rx_ghrsc: message queue resource to RX from RM
+ * @active_rx_message: ongoing gunyah_rm_message for which we're receiving fragments
+ * @call_xarray: xarray to allocate & lookup sequence IDs for Request/Response flows
+ * @next_seq: next ID to allocate (for xa_alloc_cyclic)
+ * @recv_msg: cached allocation for Rx messages
+ * @send_msg: cached allocation for Tx messages. Must hold @send_lock to manipulate.
+ * @send_lock: synchronization to allow only one request to be sent at a time
+ * @send_ready: completed when we know Tx message queue can take more messages
+ * @nh: notifier chain for clients interested in RM notification messages
+ */
+struct gunyah_rm {
+ struct device *dev;
+ struct gunyah_resource tx_ghrsc;
+ struct gunyah_resource rx_ghrsc;
+ struct gunyah_rm_message *active_rx_message;
+
+ struct xarray call_xarray;
+ u32 next_seq;
+
+ unsigned char recv_msg[GUNYAH_RM_MSGQ_MSG_SIZE];
+ unsigned char send_msg[GUNYAH_RM_MSGQ_MSG_SIZE];
+ struct mutex send_lock;
+ struct completion send_ready;
+ struct blocking_notifier_head nh;
+};
+
+/**
+ * gunyah_rm_error_remap() - Remap Gunyah resource manager errors into a Linux error code
+ * @rm_error: "Standard" return value from Gunyah resource manager
+ */
+static inline int gunyah_rm_error_remap(enum gunyah_rm_error rm_error)
+{
+ switch (rm_error) {
+ case GUNYAH_RM_ERROR_OK:
+ return 0;
+ case GUNYAH_RM_ERROR_UNIMPLEMENTED:
+ return -EOPNOTSUPP;
+ case GUNYAH_RM_ERROR_NOMEM:
+ return -ENOMEM;
+ case GUNYAH_RM_ERROR_NORESOURCE:
+ return -ENODEV;
+ case GUNYAH_RM_ERROR_DENIED:
+ return -EPERM;
+ case GUNYAH_RM_ERROR_BUSY:
+ return -EBUSY;
+ case GUNYAH_RM_ERROR_INVALID:
+ case GUNYAH_RM_ERROR_ARGUMENT_INVALID:
+ case GUNYAH_RM_ERROR_HANDLE_INVALID:
+ case GUNYAH_RM_ERROR_VALIDATE_FAILED:
+ case GUNYAH_RM_ERROR_MAP_FAILED:
+ case GUNYAH_RM_ERROR_MEM_INVALID:
+ case GUNYAH_RM_ERROR_MEM_INUSE:
+ case GUNYAH_RM_ERROR_MEM_RELEASED:
+ case GUNYAH_RM_ERROR_VMID_INVALID:
+ case GUNYAH_RM_ERROR_LOOKUP_FAILED:
+ case GUNYAH_RM_ERROR_IRQ_INVALID:
+ case GUNYAH_RM_ERROR_IRQ_INUSE:
+ case GUNYAH_RM_ERROR_IRQ_RELEASED:
+ return -EINVAL;
+ default:
+ return -EBADMSG;
+ }
+}
+
+static int gunyah_rm_init_message_payload(struct gunyah_rm_message *message,
+ const void *msg, size_t hdr_size,
+ size_t msg_size)
+{
+ const struct gunyah_rm_rpc_hdr *hdr = msg;
+ size_t max_buf_size, payload_size;
+
+ if (msg_size < hdr_size)
+ return -EINVAL;
+
+ payload_size = msg_size - hdr_size;
+
+ message->num_fragments = FIELD_GET(RM_RPC_FRAGMENTS_MASK, hdr->type);
+ message->fragments_received = 0;
+
+ /* There's not going to be any payload, no need to allocate buffer. */
+ if (!payload_size && !message->num_fragments)
+ return 0;
+
+ if (message->num_fragments > GUNYAH_RM_MAX_NUM_FRAGMENTS)
+ return -EINVAL;
+
+ max_buf_size = payload_size +
+ (message->num_fragments * GUNYAH_RM_PAYLOAD_SIZE);
+
+ message->payload = kzalloc(max_buf_size, GFP_KERNEL);
+ if (!message->payload)
+ return -ENOMEM;
+
+ memcpy(message->payload, msg + hdr_size, payload_size);
+ message->size = payload_size;
+ return 0;
+}
+
+static void gunyah_rm_abort_message(struct gunyah_rm *rm)
+{
+ switch (rm->active_rx_message->type) {
+ case RM_RPC_TYPE_REPLY:
+ rm->active_rx_message->reply.ret = -EIO;
+ complete(&rm->active_rx_message->reply.seq_done);
+ break;
+ case RM_RPC_TYPE_NOTIF:
+ fallthrough;
+ default:
+ kfree(rm->active_rx_message->payload);
+ kfree(rm->active_rx_message);
+ }
+
+ rm->active_rx_message = NULL;
+}
+
+static inline void gunyah_rm_try_complete_message(struct gunyah_rm *rm)
+{
+ struct gunyah_rm_message *message = rm->active_rx_message;
+
+ if (!message || message->fragments_received != message->num_fragments)
+ return;
+
+ switch (message->type) {
+ case RM_RPC_TYPE_REPLY:
+ complete(&message->reply.seq_done);
+ break;
+ case RM_RPC_TYPE_NOTIF:
+ blocking_notifier_call_chain(&rm->nh, message->msg_id,
+ message->payload);
+
+ kfree(message->payload);
+ kfree(message);
+ break;
+ default:
+ dev_err_ratelimited(rm->dev,
+ "Invalid message type (%u) received\n",
+ message->type);
+ gunyah_rm_abort_message(rm);
+ break;
+ }
+
+ rm->active_rx_message = NULL;
+}
+
+static void gunyah_rm_process_notif(struct gunyah_rm *rm, const void *msg,
+ size_t msg_size)
+{
+ const struct gunyah_rm_rpc_hdr *hdr = msg;
+ struct gunyah_rm_message *message;
+ int ret;
+
+ if (rm->active_rx_message) {
+ dev_err(rm->dev,
+ "Unexpected new notification, still processing an active message");
+ gunyah_rm_abort_message(rm);
+ }
+
+ message = kzalloc(sizeof(*message), GFP_KERNEL);
+ if (!message)
+ return;
+
+ message->type = RM_RPC_TYPE_NOTIF;
+ message->msg_id = le32_to_cpu(hdr->msg_id);
+
+ ret = gunyah_rm_init_message_payload(message, msg, sizeof(*hdr),
+ msg_size);
+ if (ret) {
+ dev_err(rm->dev,
+ "Failed to initialize message for notification: %d\n",
+ ret);
+ kfree(message);
+ return;
+ }
+
+ rm->active_rx_message = message;
+
+ gunyah_rm_try_complete_message(rm);
+}
+
+static void gunyah_rm_process_reply(struct gunyah_rm *rm, const void *msg,
+ size_t msg_size)
+{
+ const struct gunyah_rm_rpc_reply_hdr *reply_hdr = msg;
+ struct gunyah_rm_message *message;
+ u16 seq_id;
+
+ seq_id = le16_to_cpu(reply_hdr->hdr.seq);
+ message = xa_load(&rm->call_xarray, seq_id);
+
+ if (!message || message->msg_id != le32_to_cpu(reply_hdr->hdr.msg_id))
+ return;
+
+ if (rm->active_rx_message) {
+ dev_err(rm->dev,
+ "Unexpected new reply, still processing an active message");
+ gunyah_rm_abort_message(rm);
+ }
+
+ if (gunyah_rm_init_message_payload(message, msg, sizeof(*reply_hdr),
+ msg_size)) {
+ dev_err(rm->dev,
+ "Failed to alloc message buffer for sequence %d\n",
+ seq_id);
+ /* Send message complete and error the client. */
+ message->reply.ret = -ENOMEM;
+ complete(&message->reply.seq_done);
+ return;
+ }
+
+ message->reply.rm_error = le32_to_cpu(reply_hdr->err_code);
+ rm->active_rx_message = message;
+
+ gunyah_rm_try_complete_message(rm);
+}
+
+static void gunyah_rm_process_cont(struct gunyah_rm *rm,
+ struct gunyah_rm_message *message,
+ const void *msg, size_t msg_size)
+{
+ const struct gunyah_rm_rpc_hdr *hdr = msg;
+ size_t payload_size = msg_size - sizeof(*hdr);
+
+ if (!rm->active_rx_message)
+ return;
+
+ /*
+ * hdr->fragments and hdr->msg_id preserves the value from first reply
+ * or notif message. To detect mishandling, check it's still intact.
+ */
+ if (message->msg_id != le32_to_cpu(hdr->msg_id) ||
+ message->num_fragments !=
+ FIELD_GET(RM_RPC_FRAGMENTS_MASK, hdr->type)) {
+ gunyah_rm_abort_message(rm);
+ return;
+ }
+
+ memcpy(message->payload + message->size, msg + sizeof(*hdr),
+ payload_size);
+ message->size += payload_size;
+ message->fragments_received++;
+
+ gunyah_rm_try_complete_message(rm);
+}
+
+static irqreturn_t gunyah_rm_rx(int irq, void *data)
+{
+ enum gunyah_error gunyah_error;
+ struct gunyah_rm_rpc_hdr *hdr;
+ struct gunyah_rm *rm = data;
+ void *msg = &rm->recv_msg[0];
+ size_t len;
+ bool ready;
+
+ do {
+ gunyah_error = gunyah_hypercall_msgq_recv(rm->rx_ghrsc.capid,
+ msg,
+ sizeof(rm->recv_msg),
+ &len, &ready);
+ if (gunyah_error != GUNYAH_ERROR_OK) {
+ if (gunyah_error != GUNYAH_ERROR_MSGQUEUE_EMPTY)
+ dev_warn(rm->dev,
+ "Failed to receive data: %d\n",
+ gunyah_error);
+ return IRQ_HANDLED;
+ }
+
+ if (len < sizeof(*hdr)) {
+ dev_err_ratelimited(
+ rm->dev,
+ "Too small message received. size=%ld\n", len);
+ continue;
+ }
+
+ hdr = msg;
+ if (hdr->api != RM_RPC_API) {
+ dev_err(rm->dev, "Unknown RM RPC API version: %x\n",
+ hdr->api);
+ return IRQ_HANDLED;
+ }
+
+ switch (FIELD_GET(RM_RPC_TYPE_MASK, hdr->type)) {
+ case RM_RPC_TYPE_NOTIF:
+ gunyah_rm_process_notif(rm, msg, len);
+ break;
+ case RM_RPC_TYPE_REPLY:
+ gunyah_rm_process_reply(rm, msg, len);
+ break;
+ case RM_RPC_TYPE_CONTINUATION:
+ gunyah_rm_process_cont(rm, rm->active_rx_message, msg,
+ len);
+ break;
+ default:
+ dev_err(rm->dev,
+ "Invalid message type (%lu) received\n",
+ FIELD_GET(RM_RPC_TYPE_MASK, hdr->type));
+ return IRQ_HANDLED;
+ }
+ } while (ready);
+
+ return IRQ_HANDLED;
+}
+
+static irqreturn_t gunyah_rm_tx(int irq, void *data)
+{
+ struct gunyah_rm *rm = data;
+
+ complete_all(&rm->send_ready);
+
+ return IRQ_HANDLED;
+}
+
+static int gunyah_rm_msgq_send(struct gunyah_rm *rm, size_t size, bool push)
+ __must_hold(&rm->send_lock)
+{
+ const u64 tx_flags = push ? GUNYAH_HYPERCALL_MSGQ_TX_FLAGS_PUSH : 0;
+ enum gunyah_error gunyah_error;
+ void *data = &rm->send_msg[0];
+ bool ready;
+
+again:
+ wait_for_completion(&rm->send_ready);
+ /* reinit completion before hypercall. As soon as hypercall returns, we could get the
+ * ready interrupt. This might be before we have time to reinit the completion
+ */
+ reinit_completion(&rm->send_ready);
+ gunyah_error = gunyah_hypercall_msgq_send(rm->tx_ghrsc.capid, size,
+ data, tx_flags, &ready);
+
+ /* Should never happen because Linux properly tracks the ready-state of the msgq */
+ if (WARN_ON(gunyah_error == GUNYAH_ERROR_MSGQUEUE_FULL))
+ goto again;
+
+ if (ready)
+ complete_all(&rm->send_ready);
+
+ return gunyah_error_remap(gunyah_error);
+}
+
+static int gunyah_rm_send_request(struct gunyah_rm *rm, u32 message_id,
+ const void *req_buf, size_t req_buf_size,
+ struct gunyah_rm_message *message)
+{
+ size_t buf_size_remaining = req_buf_size;
+ const void *req_buf_curr = req_buf;
+ struct gunyah_rm_rpc_hdr *hdr =
+ (struct gunyah_rm_rpc_hdr *)&rm->send_msg[0];
+ struct gunyah_rm_rpc_hdr hdr_template;
+ void *payload = hdr + 1;
+ u32 cont_fragments = 0;
+ size_t payload_size;
+ bool push;
+ int ret;
+
+ if (req_buf_size >
+ GUNYAH_RM_MAX_NUM_FRAGMENTS * GUNYAH_RM_PAYLOAD_SIZE) {
+ dev_warn(
+ rm->dev,
+ "Limit (%lu bytes) exceeded for the maximum message size: %lu\n",
+ GUNYAH_RM_MAX_NUM_FRAGMENTS * GUNYAH_RM_PAYLOAD_SIZE,
+ req_buf_size);
+ dump_stack();
+ return -E2BIG;
+ }
+
+ if (req_buf_size)
+ cont_fragments = (req_buf_size - 1) / GUNYAH_RM_PAYLOAD_SIZE;
+
+ hdr_template.api = RM_RPC_API;
+ hdr_template.type = FIELD_PREP(RM_RPC_TYPE_MASK, RM_RPC_TYPE_REQUEST) |
+ FIELD_PREP(RM_RPC_FRAGMENTS_MASK, cont_fragments);
+ hdr_template.seq = cpu_to_le16(message->reply.seq);
+ hdr_template.msg_id = cpu_to_le32(message_id);
+
+ ret = mutex_lock_interruptible(&rm->send_lock);
+ if (ret)
+ return ret;
+
+ do {
+ *hdr = hdr_template;
+
+ /* Copy payload */
+ payload_size = min(buf_size_remaining, GUNYAH_RM_PAYLOAD_SIZE);
+ memcpy(payload, req_buf_curr, payload_size);
+ req_buf_curr += payload_size;
+ buf_size_remaining -= payload_size;
+
+ /* Only the last message should have push flag set */
+ push = !buf_size_remaining;
+ ret = gunyah_rm_msgq_send(rm, sizeof(*hdr) + payload_size,
+ push);
+ if (ret)
+ break;
+
+ hdr_template.type =
+ FIELD_PREP(RM_RPC_TYPE_MASK, RM_RPC_TYPE_CONTINUATION) |
+ FIELD_PREP(RM_RPC_FRAGMENTS_MASK, cont_fragments);
+ } while (buf_size_remaining);
+
+ mutex_unlock(&rm->send_lock);
+ return ret;
+}
+
+/**
+ * gunyah_rm_call: Achieve request-response type communication with RPC
+ * @rm: Pointer to Gunyah resource manager internal data
+ * @message_id: The RM RPC message-id
+ * @req_buf: Request buffer that contains the payload
+ * @req_buf_size: Total size of the payload
+ * @resp_buf: Pointer to a response buffer
+ * @resp_buf_size: Size of the response buffer
+ *
+ * Make a request to the Resource Manager and wait for reply back. For a successful
+ * response, the function returns the payload. The size of the payload is set in
+ * resp_buf_size. The resp_buf must be freed by the caller when 0 is returned
+ * and resp_buf_size != 0.
+ *
+ * req_buf should be not NULL for req_buf_size >0. If req_buf_size == 0,
+ * req_buf *can* be NULL and no additional payload is sent.
+ *
+ * Context: Process context. Will sleep waiting for reply.
+ * Return: 0 on success. <0 if error.
+ */
+int gunyah_rm_call(struct gunyah_rm *rm, u32 message_id, const void *req_buf,
+ size_t req_buf_size, void **resp_buf, size_t *resp_buf_size)
+{
+ struct gunyah_rm_message message = { 0 };
+ u32 seq_id;
+ int ret;
+
+ /* message_id 0 is reserved. req_buf_size implies req_buf is not NULL */
+ if (!rm || !message_id || (!req_buf && req_buf_size))
+ return -EINVAL;
+
+ message.type = RM_RPC_TYPE_REPLY;
+ message.msg_id = message_id;
+
+ message.reply.seq_done =
+ COMPLETION_INITIALIZER_ONSTACK(message.reply.seq_done);
+
+ /* Allocate a new seq number for this message */
+ ret = xa_alloc_cyclic(&rm->call_xarray, &seq_id, &message, xa_limit_16b,
+ &rm->next_seq, GFP_KERNEL);
+ if (ret < 0)
+ return ret;
+ message.reply.seq = lower_16_bits(seq_id);
+
+ /* Send the request to the Resource Manager */
+ ret = gunyah_rm_send_request(rm, message_id, req_buf, req_buf_size,
+ &message);
+ if (ret < 0) {
+ dev_warn(rm->dev, "Failed to send request. Error: %d\n", ret);
+ goto out;
+ }
+
+ /*
+ * Wait for response. Uninterruptible because rollback based on what RM did to VM
+ * requires us to know how RM handled the call.
+ */
+ wait_for_completion(&message.reply.seq_done);
+
+ /* Check for internal (kernel) error waiting for the response */
+ if (message.reply.ret) {
+ ret = message.reply.ret;
+ if (ret != -ENOMEM)
+ kfree(message.payload);
+ goto out;
+ }
+
+ /* Got a response, did resource manager give us an error? */
+ if (message.reply.rm_error != GUNYAH_RM_ERROR_OK) {
+ dev_warn(rm->dev, "RM rejected message %08x. Error: %d\n",
+ message_id, message.reply.rm_error);
+ ret = gunyah_rm_error_remap(message.reply.rm_error);
+ kfree(message.payload);
+ goto out;
+ }
+
+ /* Everything looks good, return the payload */
+ if (resp_buf_size)
+ *resp_buf_size = message.size;
+
+ if (message.size && resp_buf) {
+ *resp_buf = message.payload;
+ } else {
+ /* kfree in case RM sent us multiple fragments but never any data in
+ * those fragments. We would've allocated memory for it, but message.size == 0
+ */
+ kfree(message.payload);
+ }
+
+out:
+ xa_erase(&rm->call_xarray, message.reply.seq);
+ return ret;
+}
+
+int gunyah_rm_notifier_register(struct gunyah_rm *rm, struct notifier_block *nb)
+{
+ return blocking_notifier_chain_register(&rm->nh, nb);
+}
+EXPORT_SYMBOL_GPL(gunyah_rm_notifier_register);
+
+int gunyah_rm_notifier_unregister(struct gunyah_rm *rm,
+ struct notifier_block *nb)
+{
+ return blocking_notifier_chain_unregister(&rm->nh, nb);
+}
+EXPORT_SYMBOL_GPL(gunyah_rm_notifier_unregister);
+
+static int gunyah_platform_probe_capability(struct platform_device *pdev,
+ int idx,
+ struct gunyah_resource *ghrsc)
+{
+ int ret;
+
+ ghrsc->irq = platform_get_irq(pdev, idx);
+ if (ghrsc->irq < 0) {
+ dev_err(&pdev->dev, "Failed to get %s irq: %d\n",
+ idx ? "rx" : "tx", ghrsc->irq);
+ return ghrsc->irq;
+ }
+
+ ret = of_property_read_u64_index(pdev->dev.of_node, "reg", idx,
+ &ghrsc->capid);
+ if (ret) {
+ dev_err(&pdev->dev, "Failed to get %s capid: %d\n",
+ idx ? "rx" : "tx", ret);
+ return ret;
+ }
+
+ return 0;
+}
+
+static int gunyah_rm_probe_tx_msgq(struct gunyah_rm *rm,
+ struct platform_device *pdev)
+{
+ int ret;
+
+ rm->tx_ghrsc.type = GUNYAH_RESOURCE_TYPE_MSGQ_TX;
+ ret = gunyah_platform_probe_capability(pdev, 0, &rm->tx_ghrsc);
+ if (ret)
+ return ret;
+
+ enable_irq_wake(rm->tx_ghrsc.irq);
+
+ return devm_request_threaded_irq(rm->dev, rm->tx_ghrsc.irq, NULL,
+ gunyah_rm_tx, IRQF_ONESHOT,
+ "gunyah_rm_tx", rm);
+}
+
+static int gunyah_rm_probe_rx_msgq(struct gunyah_rm *rm,
+ struct platform_device *pdev)
+{
+ int ret;
+
+ rm->rx_ghrsc.type = GUNYAH_RESOURCE_TYPE_MSGQ_RX;
+ ret = gunyah_platform_probe_capability(pdev, 1, &rm->rx_ghrsc);
+ if (ret)
+ return ret;
+
+ enable_irq_wake(rm->rx_ghrsc.irq);
+
+ return devm_request_threaded_irq(rm->dev, rm->rx_ghrsc.irq, NULL,
+ gunyah_rm_rx, IRQF_ONESHOT,
+ "gunyah_rm_rx", rm);
+}
+
+static int gunyah_rm_probe(struct platform_device *pdev)
+{
+ struct gunyah_rm *rm;
+ int ret;
+
+ rm = devm_kzalloc(&pdev->dev, sizeof(*rm), GFP_KERNEL);
+ if (!rm)
+ return -ENOMEM;
+
+ platform_set_drvdata(pdev, rm);
+ rm->dev = &pdev->dev;
+
+ mutex_init(&rm->send_lock);
+ init_completion(&rm->send_ready);
+ BLOCKING_INIT_NOTIFIER_HEAD(&rm->nh);
+ xa_init_flags(&rm->call_xarray, XA_FLAGS_ALLOC);
+
+ ret = gunyah_rm_probe_tx_msgq(rm, pdev);
+ if (ret)
+ return ret;
+ /* assume RM is ready to receive messages from us */
+ complete_all(&rm->send_ready);
+
+ ret = gunyah_rm_probe_rx_msgq(rm, pdev);
+ if (ret)
+ return ret;
+
+ return 0;
+}
+
+static const struct of_device_id gunyah_rm_of_match[] = {
+ { .compatible = "gunyah-resource-manager" },
+ {}
+};
+MODULE_DEVICE_TABLE(of, gunyah_rm_of_match);
+
+static struct platform_driver gunyah_rm_driver = {
+ .probe = gunyah_rm_probe,
+ .driver = {
+ .name = "gunyah_rsc_mgr",
+ .of_match_table = gunyah_rm_of_match,
+ },
+};
+module_platform_driver(gunyah_rm_driver);
+
+MODULE_LICENSE("GPL");
+MODULE_DESCRIPTION("Gunyah Resource Manager Driver");
diff --git a/drivers/virt/gunyah/rsc_mgr.h b/drivers/virt/gunyah/rsc_mgr.h
new file mode 100644
index 000000000000..21318ef25040
--- /dev/null
+++ b/drivers/virt/gunyah/rsc_mgr.h
@@ -0,0 +1,28 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (c) 2022-2024 Qualcomm Innovation Center, Inc. All rights reserved.
+ */
+#ifndef __GUNYAH_RSC_MGR_PRIV_H
+#define __GUNYAH_RSC_MGR_PRIV_H
+
+#include <linux/gunyah.h>
+#include <linux/notifier.h>
+#include <linux/types.h>
+
+#define GUNYAH_VMID_INVAL U16_MAX
+
+struct gunyah_rm;
+
+int gunyah_rm_notifier_register(struct gunyah_rm *rm,
+ struct notifier_block *nb);
+int gunyah_rm_notifier_unregister(struct gunyah_rm *rm,
+ struct notifier_block *nb);
+struct device *gunyah_rm_get(struct gunyah_rm *rm);
+void gunyah_rm_put(struct gunyah_rm *rm);
+
+
+int gunyah_rm_call(struct gunyah_rm *rsc_mgr, u32 message_id,
+ const void *req_buf, size_t req_buf_size, void **resp_buf,
+ size_t *resp_buf_size);
+
+#endif
--
2.34.1
On Tue, 9 Jan 2024 11:37:45 -0800 Elliot Berman <[email protected]>
> +
> +static irqreturn_t gunyah_rm_tx(int irq, void *data)
> +{
> + struct gunyah_rm *rm = data;
> +
> + complete_all(&rm->send_ready);
> +
> + return IRQ_HANDLED;
> +}
> +
> +static int gunyah_rm_msgq_send(struct gunyah_rm *rm, size_t size, bool push)
> + __must_hold(&rm->send_lock)
> +{
> + const u64 tx_flags = push ? GUNYAH_HYPERCALL_MSGQ_TX_FLAGS_PUSH : 0;
> + enum gunyah_error gunyah_error;
> + void *data = &rm->send_msg[0];
> + bool ready;
lockdep_assert_held(&rm->send_lock); instead of __must_hold
> +
> +again:
> + wait_for_completion(&rm->send_ready);
> + /* reinit completion before hypercall. As soon as hypercall returns, we could get the
> + * ready interrupt. This might be before we have time to reinit the completion
> + */
> + reinit_completion(&rm->send_ready);
Given wait_for_completion(&rm->send_ready) with rm->send_lock held,
complete(&rm->send_ready) works for you with the bonus of cutting this
reinit off.
> + gunyah_error = gunyah_hypercall_msgq_send(rm->tx_ghrsc.capid, size,
> + data, tx_flags, &ready);
> +
> + /* Should never happen because Linux properly tracks the ready-state of the msgq */
> + if (WARN_ON(gunyah_error == GUNYAH_ERROR_MSGQUEUE_FULL))
> + goto again;
> +
> + if (ready)
> + complete_all(&rm->send_ready);
> +
> + return gunyah_error_remap(gunyah_error);
> +}
On 1/10/2024 3:41 AM, Hillf Danton wrote:
> On Tue, 9 Jan 2024 11:37:45 -0800 Elliot Berman <[email protected]>
>> +
>> +static irqreturn_t gunyah_rm_tx(int irq, void *data)
>> +{
>> + struct gunyah_rm *rm = data;
>> +
>> + complete_all(&rm->send_ready);
>> +
>> + return IRQ_HANDLED;
>> +}
>> +
>> +static int gunyah_rm_msgq_send(struct gunyah_rm *rm, size_t size, bool push)
>> + __must_hold(&rm->send_lock)
>> +{
>> + const u64 tx_flags = push ? GUNYAH_HYPERCALL_MSGQ_TX_FLAGS_PUSH : 0;
>> + enum gunyah_error gunyah_error;
>> + void *data = &rm->send_msg[0];
>> + bool ready;
>
> lockdep_assert_held(&rm->send_lock); instead of __must_hold
>> +
>> +again:
>> + wait_for_completion(&rm->send_ready);
>> + /* reinit completion before hypercall. As soon as hypercall returns, we could get the
>> + * ready interrupt. This might be before we have time to reinit the completion
>> + */
>> + reinit_completion(&rm->send_ready);
>
> Given wait_for_completion(&rm->send_ready) with rm->send_lock held,
> complete(&rm->send_ready) works for you with the bonus of cutting this
> reinit off.
>
Good point, thanks! Applied both suggestions
- Elliot