add descriptor table for mapping fds to handles (#464)

* add descriptor table for mapping fds to handles

This introduces `descriptor_table.h` and `descriptor_table.c`, providing a
global hashtable for tracking `wasi-libc`-managed file descriptors.

WASI Preview 2 has no notion of file descriptors and instead uses unforgeable
resource handles.  Moreover, there's not necessarily a one-to-one correspondence
between POSIX file descriptors and resource handles (e.g. a TCP connection may
require separate handles for reading, writing, and polling the same connection).
We use this table to map each POSIX descriptor to a set of one or more handles
and any extra state which libc needs to track.

Note that we've added `descriptor_table.h` to the
libc-bottom-half/headers/public/wasi directory, making it part of the public
API.  The intention is to give applications access to the mapping, enabling them
to convert descriptors to handles and vice-versa should they need to
interoperate with both libc and WASI directly.

Co-authored-by: Dave Bakker <github@davebakker.io>
Signed-off-by: Joel Dice <joel.dice@fermyon.com>

* add dummy fields to otherwise empty structs

The C standard doesn't allow empty structs.  Clang doesn't currently complain,
but we might as well stick to the spec in case it becomes more strict in the
future.

Signed-off-by: Joel Dice <joel.dice@fermyon.com>

* move descriptor_table.h to headers/private

We're not yet ready to commit to making this API public, so we'll make it
private for now.

I've also expanded a comment in descriptor_table.c to explain the current ABI
for resource handles.

Signed-off-by: Joel Dice <joel.dice@fermyon.com>

* re-run clang-format to fix indentation

Signed-off-by: Joel Dice <joel.dice@fermyon.com>

---------

Signed-off-by: Joel Dice <joel.dice@fermyon.com>
Co-authored-by: Dave Bakker <github@davebakker.io>

Author: dicej

Makefile

@@ -79,10 +79,12 @@ ifeq ($(WASI_SNAPSHOT), preview1)
 # Omit source files not relevant to WASI Preview 1.  As we introduce files
 # supporting `wasi-sockets` for `wasm32-wasi-preview2`, we'll add those files to
 # this list.
-LIBC_BOTTOM_HALF_OMIT_SOURCES := $(LIBC_BOTTOM_HALF_SOURCES)/preview2.c
+LIBC_BOTTOM_HALF_OMIT_SOURCES := \
+	$(LIBC_BOTTOM_HALF_SOURCES)/preview2.c \
+	$(LIBC_BOTTOM_HALF_SOURCES)/descriptor_table.c
 LIBC_BOTTOM_HALF_ALL_SOURCES := $(filter-out $(LIBC_BOTTOM_HALF_OMIT_SOURCES),$(LIBC_BOTTOM_HALF_ALL_SOURCES))
-# Omit preview2.h from include-all.c test.
-INCLUDE_ALL_CLAUSES := -not -name preview2.h
+# Omit preview2-specific headers from include-all.c test.
+INCLUDE_ALL_CLAUSES := -not -name preview2.h -not -name descriptor_table.h
 endif
 
 # FIXME(https://reviews.llvm.org/D85567) - due to a bug in LLD the weak

expected/wasm32-wasi-preview2/defined-symbols.txt

@@ -509,6 +509,9 @@ ctanhl
 ctanl
 ctime
 ctime_r
+descriptor_table_get_ref
+descriptor_table_insert
+descriptor_table_remove
 difftime
 dirfd
 dirname

expected/wasm32-wasi-preview2/undefined-symbols.txt

@@ -65,6 +65,8 @@ __subtf3
 __trunctfdf2
 __trunctfsf2
 __unordtf2
+__wasi_preview1_adapter_close_badfd
+__wasi_preview1_adapter_open_badfd
 __wasm_call_ctors
 __wasm_import_environment_get_arguments
 __wasm_import_environment_get_environment

libc-bottom-half/headers/private/wasi/descriptor_table.h

@@ -0,0 +1,127 @@
+#ifndef DESCRIPTOR_TABLE_H
+#define DESCRIPTOR_TABLE_H
+
+#include <wasi/preview2.h>
+
+typedef struct {
+	int dummy;
+} tcp_socket_state_unbound_t;
+typedef struct {
+	int dummy;
+} tcp_socket_state_bound_t;
+typedef struct {
+	int dummy;
+} tcp_socket_state_connecting_t;
+typedef struct {
+	int dummy;
+} tcp_socket_state_listening_t;
+
+typedef struct {
+	streams_own_input_stream_t input;
+	poll_own_pollable_t input_pollable;
+	streams_own_output_stream_t output;
+	poll_own_pollable_t output_pollable;
+} tcp_socket_state_connected_t;
+
+typedef struct {
+	network_error_code_t error_code;
+} tcp_socket_state_connect_failed_t;
+
+// This is a tagged union. When adding/removing/renaming cases, be sure to keep the tag and union definitions in sync.
+typedef struct {
+	enum {
+		TCP_SOCKET_STATE_UNBOUND,
+		TCP_SOCKET_STATE_BOUND,
+		TCP_SOCKET_STATE_CONNECTING,
+		TCP_SOCKET_STATE_CONNECTED,
+		TCP_SOCKET_STATE_CONNECT_FAILED,
+		TCP_SOCKET_STATE_LISTENING,
+	} tag;
+	union {
+		tcp_socket_state_unbound_t unbound;
+		tcp_socket_state_bound_t bound;
+		tcp_socket_state_connecting_t connecting;
+		tcp_socket_state_connected_t connected;
+		tcp_socket_state_connect_failed_t connect_failed;
+		tcp_socket_state_listening_t listening;
+	};
+} tcp_socket_state_t;
+
+typedef struct {
+	tcp_own_tcp_socket_t socket;
+	poll_own_pollable_t socket_pollable;
+	bool blocking;
+	bool fake_nodelay;
+	bool fake_reuseaddr;
+	network_ip_address_family_t family;
+	tcp_socket_state_t state;
+} tcp_socket_t;
+
+typedef struct {
+	udp_own_incoming_datagram_stream_t incoming;
+	poll_own_pollable_t incoming_pollable;
+	udp_own_outgoing_datagram_stream_t outgoing;
+	poll_own_pollable_t outgoing_pollable;
+} udp_socket_streams_t;
+
+typedef struct {
+	int dummy;
+} udp_socket_state_unbound_t;
+typedef struct {
+	int dummy;
+} udp_socket_state_bound_nostreams_t;
+
+typedef struct {
+	udp_socket_streams_t streams; // Streams have no remote_address
+} udp_socket_state_bound_streaming_t;
+
+typedef struct {
+	udp_socket_streams_t streams; // Streams have a remote_address
+} udp_socket_state_connected_t;
+
+// This is a tagged union. When adding/removing/renaming cases, be sure to keep the tag and union definitions in sync.
+// The "bound" state is split up into two distinct tags:
+// - "bound_nostreams": Bound, but no datagram streams set up (yet). That will be done the first time send or recv is called.
+// - "bound_streaming": Bound with active streams.
+typedef struct {
+	enum {
+		UDP_SOCKET_STATE_UNBOUND,
+		UDP_SOCKET_STATE_BOUND_NOSTREAMS,
+		UDP_SOCKET_STATE_BOUND_STREAMING,
+		UDP_SOCKET_STATE_CONNECTED,
+	} tag;
+	union {
+		udp_socket_state_unbound_t unbound;
+		udp_socket_state_bound_nostreams_t bound_nostreams;
+		udp_socket_state_bound_streaming_t bound_streaming;
+		udp_socket_state_connected_t connected;
+	};
+} udp_socket_state_t;
+
+typedef struct {
+	udp_own_udp_socket_t socket;
+	poll_own_pollable_t socket_pollable;
+	bool blocking;
+	network_ip_address_family_t family;
+	udp_socket_state_t state;
+} udp_socket_t;
+
+// This is a tagged union. When adding/removing/renaming cases, be sure to keep the tag and union definitions in sync.
+typedef struct {
+	enum {
+		DESCRIPTOR_TABLE_ENTRY_TCP_SOCKET,
+		DESCRIPTOR_TABLE_ENTRY_UDP_SOCKET,
+	} tag;
+	union {
+		tcp_socket_t tcp_socket;
+		udp_socket_t udp_socket;
+	};
+} descriptor_table_entry_t;
+
+bool descriptor_table_insert(descriptor_table_entry_t entry, int *fd);
+
+bool descriptor_table_get_ref(int fd, descriptor_table_entry_t **entry);
+
+bool descriptor_table_remove(int fd, descriptor_table_entry_t *entry);
+
+#endif