/*
 * Copyright 2021 Google LLC
 * SPDX-License-Identifier: MIT
 */

#include "render_context.h"

#include <sys/mman.h>

#include "util/u_thread.h"
#include "virgl_util.h"

#include "render_state.h"

void
render_context_update_timeline(struct render_context *ctx,
                               uint32_t ring_idx,
                               uint32_t seqno)
{
   /* this can be called by the context's main thread and sync threads */
   atomic_store(&ctx->shmem_timelines[ring_idx], seqno);
   if (ctx->fence_eventfd >= 0)
      write_eventfd(ctx->fence_eventfd, 1);
}

static bool
render_context_dispatch_submit_fence(struct render_context *ctx,
                                     const union render_context_op_request *request,
                                     UNUSED const int *fds,
                                     UNUSED int fd_count)
{
   const struct render_context_op_submit_fence_request *req = &request->submit_fence;

   /* always merge fences */
   assert(!(req->flags & ~VIRGL_RENDERER_FENCE_FLAG_MERGEABLE));
   assert(req->ring_index < (uint32_t)ctx->timeline_count);
   return render_state_submit_fence(ctx->ctx_id, VIRGL_RENDERER_FENCE_FLAG_MERGEABLE,
                                    req->ring_index, req->seqno);
}

static bool
render_context_dispatch_submit_cmd(struct render_context *ctx,
                                   const union render_context_op_request *request,
                                   UNUSED const int *fds,
                                   UNUSED int fd_count)
{
   const struct render_context_op_submit_cmd_request *req = &request->submit_cmd;
   void *cmd = (void *)req->cmd;
   if (req->size > sizeof(req->cmd)) {
      cmd = malloc(req->size);
      if (!cmd)
         return true;

      const size_t inlined = sizeof(req->cmd);
      const size_t remain = req->size - inlined;

      memcpy(cmd, req->cmd, inlined);
      if (!render_socket_receive_data(&ctx->socket, (char *)cmd + inlined, remain)) {
         free(cmd);
         return false;
      }
   }

   bool ok = render_state_submit_cmd(ctx->ctx_id, cmd, req->size);

   if (cmd != req->cmd)
      free(cmd);

   return ok;
}

static bool
render_context_dispatch_destroy_resource(struct render_context *ctx,
                                         const union render_context_op_request *req,
                                         UNUSED const int *fds,
                                         UNUSED int fd_count)
{
   render_state_destroy_resource(ctx->ctx_id, req->destroy_resource.res_id);
   return true;
}

static bool
render_context_dispatch_import_resource(struct render_context *ctx,
                                        const union render_context_op_request *request,
                                        const int *fds,
                                        int fd_count)
{
   if (fd_count != 1) {
      render_log("failed to attach resource with fd_count %d", fd_count);
      return false;
   }

   /* classic 3d resource with valid size reuses the blob import path here */
   const struct render_context_op_import_resource_request *req =
      &request->import_resource;
   return render_state_import_resource(ctx->ctx_id, req->res_id, req->fd_type, fds[0],
                                       req->size);
}

static bool
render_context_dispatch_create_resource(struct render_context *ctx,
                                        const union render_context_op_request *request,
                                        UNUSED const int *fds,
                                        UNUSED int fd_count)
{
   const struct render_context_op_create_resource_request *req =
      &request->create_resource;
   struct render_context_op_create_resource_reply reply = {
      .fd_type = VIRGL_RESOURCE_FD_INVALID,
   };
   int res_fd;
   bool ok = render_state_create_resource(ctx->ctx_id, req->res_id, req->blob_id,
                                          req->blob_size, req->blob_flags, &reply.fd_type,
                                          &res_fd, &reply.map_info, &reply.vulkan_info);
   if (!ok)
      return render_socket_send_reply(&ctx->socket, &reply, sizeof(reply));

   ok =
      render_socket_send_reply_with_fds(&ctx->socket, &reply, sizeof(reply), &res_fd, 1);
   close(res_fd);

   return ok;
}

static bool
render_context_dispatch_init(struct render_context *ctx,
                             const union render_context_op_request *request,
                             const int *fds,
                             int fd_count)
{
   if (fd_count != 1 && fd_count != 2)
      return false;

   const struct render_context_op_init_request *req = &request->init;
   const int timeline_count = req->shmem_size / sizeof(*ctx->shmem_timelines);
   const int shmem_fd = fds[0];
   const int fence_eventfd = fd_count == 2 ? fds[1] : -1;

   void *shmem_ptr = mmap(NULL, req->shmem_size, PROT_WRITE, MAP_SHARED, shmem_fd, 0);
   if (shmem_ptr == MAP_FAILED)
      return false;

   if (!render_state_create_context(ctx, req->flags, ctx->name_len, ctx->name)) {
      munmap(shmem_ptr, req->shmem_size);
      return false;
   }

   ctx->shmem_fd = shmem_fd;
   ctx->shmem_size = req->shmem_size;
   ctx->shmem_ptr = shmem_ptr;
   ctx->shmem_timelines = shmem_ptr;

   for (int i = 0; i < timeline_count; i++)
      atomic_store(&ctx->shmem_timelines[i], 0);

   ctx->timeline_count = timeline_count;

   ctx->fence_eventfd = fence_eventfd;

   return true;
}

static bool
render_context_dispatch_nop(UNUSED struct render_context *ctx,
                            UNUSED const union render_context_op_request *req,
                            UNUSED const int *fds,
                            UNUSED int fd_count)
{
   return true;
}

struct render_context_dispatch_entry {
   size_t expect_size;
   int max_fd_count;
   bool (*dispatch)(struct render_context *ctx,
                    const union render_context_op_request *req,
                    const int *fds,
                    int fd_count);
};

static const struct render_context_dispatch_entry
   render_context_dispatch_table[RENDER_CONTEXT_OP_COUNT] = {
#define RENDER_CONTEXT_DISPATCH(NAME, name, max_fd)                                      \
   [RENDER_CONTEXT_OP_##                                                                 \
      NAME] = { .expect_size = sizeof(struct render_context_op_##name##_request),        \
                .max_fd_count = (max_fd),                                                \
                .dispatch = render_context_dispatch_##name }
      RENDER_CONTEXT_DISPATCH(NOP, nop, 0),
      RENDER_CONTEXT_DISPATCH(INIT, init, 2),
      RENDER_CONTEXT_DISPATCH(CREATE_RESOURCE, create_resource, 0),
      RENDER_CONTEXT_DISPATCH(IMPORT_RESOURCE, import_resource, 1),
      RENDER_CONTEXT_DISPATCH(DESTROY_RESOURCE, destroy_resource, 0),
      RENDER_CONTEXT_DISPATCH(SUBMIT_CMD, submit_cmd, 0),
      RENDER_CONTEXT_DISPATCH(SUBMIT_FENCE, submit_fence, 0),
#undef RENDER_CONTEXT_DISPATCH
   };

static bool
render_context_dispatch(struct render_context *ctx)
{
   union render_context_op_request req;
   size_t req_size;
   int req_fds[8];
   int req_fd_count;
   if (!render_socket_receive_request_with_fds(&ctx->socket, &req, sizeof(req), &req_size,
                                               req_fds, ARRAY_SIZE(req_fds),
                                               &req_fd_count))
      return false;

   assert((unsigned int)req_fd_count <= ARRAY_SIZE(req_fds));

   if (req.header.op >= RENDER_CONTEXT_OP_COUNT) {
      render_log("invalid context op %d", req.header.op);
      goto fail;
   }

   const struct render_context_dispatch_entry *entry =
      &render_context_dispatch_table[req.header.op];
   if (entry->expect_size != req_size || entry->max_fd_count < req_fd_count) {
      render_log("invalid request size (%zu) or fd count (%d) for context op %d",
                 req_size, req_fd_count, req.header.op);
      goto fail;
   }

   const bool ok = entry->dispatch(ctx, &req, req_fds, req_fd_count);
   if (!ok) {
      render_log("failed to dispatch context op %d", req.header.op);
      goto fail;
   }

   return true;

fail:
   for (int i = 0; i < req_fd_count; i++)
      close(req_fds[i]);
   return false;
}

static bool
render_context_run(struct render_context *ctx)
{
   while (true) {
      if (!render_context_dispatch(ctx))
         return false;
   }

   return true;
}

static void
render_context_fini(struct render_context *ctx)
{
   /* destroy the context first to join its sync threads and ring threads */
   render_state_destroy_context(ctx->ctx_id);

   if (ctx->shmem_ptr)
      munmap(ctx->shmem_ptr, ctx->shmem_size);
   if (ctx->shmem_fd >= 0)
      close(ctx->shmem_fd);

   if (ctx->fence_eventfd >= 0)
      close(ctx->fence_eventfd);

   if (ctx->name)
      free(ctx->name);

   render_socket_fini(&ctx->socket);
}

static void
render_context_set_thread_name(uint32_t ctx_id, UNUSED const char *ctx_name)
{
   char thread_name[16];

#pragma GCC diagnostic push
#if !defined(__clang__) || __clang_major__ >= 18
#pragma GCC diagnostic ignored "-Wformat-truncation"
#endif
   /* context name may match guest process name, so use a generic name in
    * production/release builds.
    */
#ifndef NDEBUG
   snprintf(thread_name, ARRAY_SIZE(thread_name), "virgl-%d-%s", ctx_id, ctx_name);
#else
   snprintf(thread_name, ARRAY_SIZE(thread_name), "virgl-%d-gpu_renderer", ctx_id);
#endif
#pragma GCC diagnostic pop

   u_thread_setname(thread_name);
}

static const char *ctx_name_expansions[] = {
   "DOOMEternalx64vk.exe",
};

static bool
render_context_init_name(struct render_context *ctx,
                         uint32_t ctx_id,
                         const char *ctx_name)
{
   /* Linux guests may only pass the first 15 chars of a guest application name,
    * plus a null terminator. In that case, attempt substring matching and
    * expansion to enable proper dri-conf lookup in native mesa drivers.
    */
   static const size_t truncated_name_len = 15;
   size_t name_len = strlen(ctx_name);
   if (name_len == truncated_name_len) {
      for (uint32_t i = 0; i < ARRAY_SIZE(ctx_name_expansions); i++) {
         const char *full = ctx_name_expansions[i];
         if (!strncmp(ctx_name, full, truncated_name_len)) {
            ctx_name = full;
            name_len = strlen(full);
            break;
         }
      }
   }

   ctx->name_len = name_len;
   ctx->name = malloc(name_len + 1);
   if (!ctx->name)
      return false;

   strcpy(ctx->name, ctx_name);

   /* Overrides the executable name used only by mesa to load app-specific
    * driver configuration.
    */
   setenv("MESA_DRICONF_EXECUTABLE_OVERRIDE", ctx->name, 0);

   /* Host Mesa still sees the process name as "virgl_render_server" unless
    * additionally overridden by setenv("MESA_PROCESS_NAME", ...).
    */
   render_context_set_thread_name(ctx_id, ctx->name);

   return true;
}

static bool
render_context_init(struct render_context *ctx, const struct render_context_args *args)
{
   memset(ctx, 0, sizeof(*ctx));
   ctx->ctx_id = args->ctx_id;
   render_socket_init(&ctx->socket, args->ctx_fd);
   ctx->shmem_fd = -1;
   ctx->fence_eventfd = -1;

   if (!render_context_init_name(ctx, args->ctx_id, args->ctx_name))
      return false;

   return true;
}

bool
render_context_main(const struct render_context_args *args)
{
   struct render_context ctx;

   assert(args->valid && args->ctx_id && args->ctx_fd >= 0);

   if (!render_state_init(args->init_flags)) {
      close(args->ctx_fd);
      return false;
   }

   if (!render_context_init(&ctx, args)) {
      render_state_fini();
      close(args->ctx_fd);
      return false;
   }

   const bool ok = render_context_run(&ctx);
   render_context_fini(&ctx);

   render_state_fini();

   return ok;
}