Commit f40fdc87 authored by Marek Vavruša's avatar Marek Vavruša

daemon/io: don’t use handle data for buffers

TODO: per-loop freelists are implemented
parent 970a12c7
......@@ -23,31 +23,21 @@
#define ENDPOINT_BUFSIZE 512 /**< This is an artificial limit for DNS query. */
static void *buf_alloc(void)
{
struct endpoint_data *data = malloc(sizeof(*data) + ENDPOINT_BUFSIZE);
if (data == NULL) {
return NULL;
}
data->buf = uv_buf_init((char *)data + sizeof(*data), ENDPOINT_BUFSIZE);
return data;
}
static void buf_get(uv_handle_t* handle, size_t suggested_size, uv_buf_t* buf)
{
struct endpoint_data *data = handle->data;
*buf = data->buf;
}
static void buf_free(uv_handle_t* handle)
{
free(handle->data);
#warning TODO: freelist from worker allocation
buf->base = malloc(ENDPOINT_BUFSIZE);
if (buf->base) {
buf->len = ENDPOINT_BUFSIZE;
} else {
buf->len = 0;
}
}
static void udp_send(uv_udp_t *handle, knot_pkt_t *answer, const struct sockaddr *addr)
int udp_send(uv_udp_t *handle, knot_pkt_t *answer, const struct sockaddr *addr)
{
uv_buf_t sendbuf = uv_buf_init((char *)answer->wire, answer->size);
uv_udp_try_send(handle, &sendbuf, 1, addr);
return uv_udp_try_send(handle, &sendbuf, 1, addr);
}
static void udp_recv(uv_udp_t *handle, ssize_t nread, const uv_buf_t *buf,
......@@ -74,6 +64,19 @@ static void udp_recv(uv_udp_t *handle, ssize_t nread, const uv_buf_t *buf,
/* Cleanup */
knot_pkt_free(&query);
knot_pkt_free(&answer);
free(buf->base);
}
static uv_udp_t *udp_create(uv_loop_t *loop)
{
uv_udp_t *handle = malloc(sizeof(uv_udp_t));
if (!handle) {
return handle;
}
uv_udp_init(loop, handle);
return handle;
}
int udp_bind(struct endpoint *ep, struct sockaddr *addr)
......@@ -84,28 +87,19 @@ int udp_bind(struct endpoint *ep, struct sockaddr *addr)
return ret;
}
handle->data = buf_alloc();
if (handle->data == NULL) {
udp_unbind(ep);
return kr_error(ENOMEM);
}
uv_udp_recv_start(handle, &buf_get, &udp_recv);
return 0;
return uv_udp_recv_start(handle, &buf_get, &udp_recv);
}
void udp_unbind(struct endpoint *ep)
{
uv_udp_t *handle = &ep->udp;
uv_udp_recv_stop(handle);
buf_free((uv_handle_t *)handle);
uv_close((uv_handle_t *)handle, NULL);
}
static void tcp_unbind_handle(uv_handle_t *handle)
{
uv_read_stop((uv_stream_t *)handle);
buf_free(handle);
}
static void tcp_send(uv_handle_t *handle, const knot_pkt_t *answer)
......@@ -148,28 +142,28 @@ static void tcp_recv(uv_stream_t *handle, ssize_t nread, const uv_buf_t *buf)
/* Cleanup */
knot_pkt_free(&query);
knot_pkt_free(&answer);
free(buf->base);
}
static void tcp_accept(uv_stream_t *master, int status)
static uv_tcp_t *tcp_create(uv_loop_t *loop)
{
if (status != 0) {
return;
uv_tcp_t *handle = malloc(sizeof(uv_tcp_t));
if (!handle) {
return handle;
}
uv_tcp_t *client = malloc(sizeof(uv_tcp_t));
if (client == NULL) {
return;
}
uv_tcp_init(loop, handle);
return handle;
}
uv_tcp_init(master->loop, client);
client->data = buf_alloc();
if (client->data == NULL) {
free(client);
static void tcp_accept(uv_stream_t *master, int status)
{
if (status != 0) {
return;
}
if (uv_accept(master, (uv_stream_t*)client) != 0) {
buf_free((uv_handle_t *)client);
uv_tcp_t *client = tcp_create(master->loop);
if (!client || uv_accept(master, (uv_stream_t*)client) != 0) {
free(client);
return;
}
......@@ -191,12 +185,6 @@ int tcp_bind(struct endpoint *ep, struct sockaddr *addr)
return ret;
}
handle->data = buf_alloc();
if (handle->data == NULL) {
tcp_unbind(ep);
return kr_error(ENOMEM);
}
return 0;
}
......
......@@ -27,10 +27,6 @@ enum endpoint_flag {
NET_TCP = 1 << 1
};
struct endpoint_data {
uv_buf_t buf;
};
struct endpoint {
uv_udp_t udp;
uv_tcp_t tcp;
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment