ovn: Add address set support.

Update the OVN expression parser to support address sets.  Previously,
you could have a set of IP or MAC addresses in this form:

    {addr1, addr2, ..., addrN}

This patch adds support for a bit of indirection where we can define a
set of addresses and refer to them by name.

    $name

This '$name' can be used in the expresssions like

    {addr1, addr2, $name, ... }
    {$name}
    $name

A future patch will expose the ability to define address sets for use.

Signed-off-by: Russell Bryant <russell@ovn.org>
Co-authored-by: Babu Shanmugam <bschanmu@redhat.com>
Signed-off-by: Babu Shanmugam <bschanmu@redhat.com>
[blp@ovn.org made numerous small changes]
Signed-off-by: Ben Pfaff <blp@ovn.org>

ovn/controller/lflow.c

@@ -306,7 +306,7 @@ add_logical_flows(struct controller_ctx *ctx, const struct lport_index *lports,
         struct hmap matches;
         struct expr *expr;
 
-        expr = expr_parse_string(lflow->match, &symtab, &error);
+        expr = expr_parse_string(lflow->match, &symtab, NULL, &error);
         if (!error) {
             if (prereqs) {
                 expr = expr_combine(EXPR_T_AND, expr, prereqs);

ovn/lib/actions.c

@@ -208,7 +208,7 @@ add_prerequisite(struct action_context *ctx, const char *prerequisite)
     struct expr *expr;
     char *error;
 
-    expr = expr_parse_string(prerequisite, ctx->ap->symtab, &error);
+    expr = expr_parse_string(prerequisite, ctx->ap->symtab, NULL, &error);
     ovs_assert(!error);
     ctx->prereqs = expr_combine(EXPR_T_AND, ctx->prereqs, expr);
 }

ovn/lib/expr.c

@@ -409,6 +409,7 @@ expr_print(const struct expr *e)
 struct expr_context {
     struct lexer *lexer;        /* Lexer for pulling more tokens. */
     const struct shash *symtab; /* Symbol table. */
+    const struct shash *macros; /* Table of macros. */
     char *error;                /* Error, if any, otherwise NULL. */
     bool not;                   /* True inside odd number of NOT operators. */
 };
@@ -728,6 +729,33 @@ assign_constant_set_type(struct expr_context *ctx,
     }
 }
 
+static bool
+parse_macros(struct expr_context *ctx, struct expr_constant_set *cs,
+             size_t *allocated_values)
+{
+    struct expr_constant_set *addr_set
+        = shash_find_data(ctx->macros, ctx->lexer->token.s);
+    if (!addr_set) {
+        expr_syntax_error(ctx, "expecting address set name.");
+        return false;
+    }
+
+    if (!assign_constant_set_type(ctx, cs, EXPR_C_INTEGER)) {
+        return false;
+    }
+
+    size_t n_values = cs->n_values + addr_set->n_values;
+    if (n_values >= *allocated_values) {
+        cs->values = xrealloc(cs->values, n_values * sizeof *cs->values);
+        *allocated_values = n_values;
+    }
+    for (size_t i = 0; i < addr_set->n_values; i++) {
+        cs->values[cs->n_values++] = addr_set->values[i];
+    }
+
+    return true;
+}
+
 static bool
 parse_constant(struct expr_context *ctx, struct expr_constant_set *cs,
                size_t *allocated_values)
@@ -759,6 +787,12 @@ parse_constant(struct expr_context *ctx, struct expr_constant_set *cs,
         }
         lexer_get(ctx->lexer);
         return true;
+    } else if (ctx->lexer->token.type == LEX_T_MACRO) {
+        if (!parse_macros(ctx, cs, allocated_values)) {
+            return false;
+        }
+        lexer_get(ctx->lexer);
+        return true;
     } else {
         expr_syntax_error(ctx, "expecting constant.");
         return false;
@@ -808,6 +842,69 @@ expr_constant_set_destroy(struct expr_constant_set *cs)
     }
 }
 
+/* Adds a macro named 'name' to 'macros', replacing any existing macro with the
+ * given name. */
+void
+expr_macros_add(struct shash *macros, const char *name,
+                const char *const *values, size_t n_values)
+{
+    /* Replace any existing entry for this name. */
+    expr_macros_remove(macros, name);
+
+    struct expr_constant_set *cs = xzalloc(sizeof *cs);
+    cs->type = EXPR_C_INTEGER;
+    cs->in_curlies = true;
+    cs->n_values = 0;
+    cs->values = xmalloc(n_values * sizeof *cs->values);
+    for (size_t i = 0; i < n_values; i++) {
+        /* Use the lexer to convert each macro into the proper
+         * integer format. */
+        struct lexer lex;
+        lexer_init(&lex, values[i]);
+        lexer_get(&lex);
+        if (lex.token.type != LEX_T_INTEGER
+            && lex.token.type != LEX_T_MASKED_INTEGER) {
+            VLOG_WARN("Invalid address set entry: '%s', token type: %d",
+                      values[i], lex.token.type);
+        } else {
+            union expr_constant *c = &cs->values[cs->n_values++];
+            c->value = lex.token.value;
+            c->format = lex.token.format;
+            c->masked = lex.token.type == LEX_T_MASKED_INTEGER;
+            if (c->masked) {
+                c->mask = lex.token.mask;
+            }
+        }
+        lexer_destroy(&lex);
+    }
+
+    shash_add(macros, name, cs);
+}
+
+void
+expr_macros_remove(struct shash *macros, const char *name)
+{
+    struct expr_constant_set *cs = shash_find_and_delete(macros, name);
+    if (cs) {
+        expr_constant_set_destroy(cs);
+        free(cs);
+    }
+}
+
+/* Destroy all contents of 'macros'. */
+void
+expr_macros_destroy(struct shash *macros)
+{
+    struct shash_node *node, *next;
+
+    SHASH_FOR_EACH_SAFE (node, next, macros) {
+        struct expr_constant_set *cs = node->data;
+
+        shash_delete(macros, node);
+        expr_constant_set_destroy(cs);
+    }
+}
+
 static struct expr *
 expr_parse_primary(struct expr_context *ctx, bool *atomic)
 {
@@ -980,9 +1077,12 @@ expr_parse__(struct expr_context *ctx)
  * The caller must eventually free the returned expression (with
  * expr_destroy()) or error (with free()). */
 struct expr *
-expr_parse(struct lexer *lexer, const struct shash *symtab, char **errorp)
+expr_parse(struct lexer *lexer, const struct shash *symtab,
+           const struct shash *macros, char **errorp)
 {
-    struct expr_context ctx = { .lexer = lexer, .symtab = symtab };
+    struct expr_context ctx = { .lexer = lexer,
+                                .symtab = symtab,
+                                .macros = macros };
     struct expr *e = expr_parse__(&ctx);
     *errorp = ctx.error;
     ovs_assert((ctx.error != NULL) != (e != NULL));
@@ -991,14 +1091,15 @@ expr_parse(struct lexer *lexer, const struct shash *symtab, char **errorp)
 
 /* Like expr_parse(), but the expression is taken from 's'. */
 struct expr *
-expr_parse_string(const char *s, const struct shash *symtab, char **errorp)
+expr_parse_string(const char *s, const struct shash *symtab,
+                  const struct shash *macros, char **errorp)
 {
     struct lexer lexer;
     struct expr *expr;
 
     lexer_init(&lexer, s);
     lexer_get(&lexer);
-    expr = expr_parse(&lexer, symtab, errorp);
+    expr = expr_parse(&lexer, symtab, macros, errorp);
     if (!*errorp && lexer.token.type != LEX_T_END) {
         *errorp = xstrdup("Extra tokens at end of input.");
         expr_destroy(expr);
@@ -1149,7 +1250,7 @@ expr_get_level(const struct expr *expr)
 static enum expr_level
 expr_parse_level(const char *s, const struct shash *symtab, char **errorp)
 {
-    struct expr *expr = expr_parse_string(s, symtab, errorp);
+    struct expr *expr = expr_parse_string(s, symtab, NULL, errorp);
     enum expr_level level = expr ? expr_get_level(expr) : EXPR_L_NOMINAL;
     expr_destroy(expr);
     return level;
@@ -1292,7 +1393,7 @@ parse_and_annotate(const char *s, const struct shash *symtab,
     char *error;
     struct expr *expr;
 
-    expr = expr_parse_string(s, symtab, &error);
+    expr = expr_parse_string(s, symtab, NULL, &error);
     if (expr) {
         expr = expr_annotate__(expr, symtab, nesting, &error);
     }

ovn/lib/expr.h

@@ -352,8 +352,10 @@ expr_from_node(const struct ovs_list *node)
 void expr_format(const struct expr *, struct ds *);
 void expr_print(const struct expr *);
 struct expr *expr_parse(struct lexer *, const struct shash *symtab,
+                        const struct shash *macros,
                         char **errorp);
 struct expr *expr_parse_string(const char *, const struct shash *symtab,
+                               const struct shash *macros,
                                char **errorp);
 
 struct expr *expr_clone(struct expr *);
@@ -453,4 +455,20 @@ char *expr_parse_constant_set(struct lexer *, const struct shash *symtab,
     OVS_WARN_UNUSED_RESULT;
 void expr_constant_set_destroy(struct expr_constant_set *cs);
 
+
+/* Address sets, aka "macros".
+ *
+ * Instead of referring to a set of value as:
+ *    {addr1, addr2, ..., addrN}
+ * You can register a set of values and refer to them as:
+ *    $name
+ * The macros should all have integer/masked-integer values.
+ * The values that don't qualify are ignored.
+ */
+
+void expr_macros_add(struct shash *macros, const char *name,
+                     const char * const *values, size_t n_values);
+void expr_macros_remove(struct shash *macros, const char *name);
+void expr_macros_destroy(struct shash *macros);
+
 #endif /* ovn/expr.h */

ovn/lib/lex.c

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2015 Nicira, Inc.
+ * Copyright (c) 2015, 2016 Nicira, Inc.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -227,6 +227,10 @@ lex_token_format(const struct lex_token *token, struct ds *s)
         lex_token_format_masked_integer(token, s);
         break;
 
+    case LEX_T_MACRO:
+        ds_put_format(s, "$%s", token->s);
+        break;
+
     case LEX_T_LPAREN:
         ds_put_cstr(s, "(");
         break;
@@ -514,7 +518,7 @@ lex_is_idn(unsigned char c)
 }
 
 static const char *
-lex_parse_id(const char *p, struct lex_token *token)
+lex_parse_id(const char *p, enum lex_type type, struct lex_token *token)
 {
     const char *start = p;
 
@@ -522,11 +526,23 @@ lex_parse_id(const char *p, struct lex_token *token)
         p++;
     } while (lex_is_idn(*p));
 
-    token->type = LEX_T_ID;
+    token->type = type;
     lex_token_strcpy(token, start, p - start);
     return p;
 }
 
+static const char *
+lex_parse_macro(const char *p, struct lex_token *token)
+{
+    p++;
+    if (!lex_is_id1(*p)) {
+        lex_error(token, "`$' must be followed by a valid identifier.");
+        return p;
+    }
+
+    return lex_parse_id(p, LEX_T_MACRO, token);
+}
+
 /* Initializes 'token' and parses the first token from the beginning of
  * null-terminated string 'p' into 'token'.  Stores a pointer to the start of
  * the token (after skipping white space and comments, if any) into '*startp'.
@@ -697,6 +713,10 @@ next:
         }
         break;
 
+    case '$':
+        p = lex_parse_macro(p, token);
+        break;
+
     case '0': case '1': case '2': case '3': case '4':
     case '5': case '6': case '7': case '8': case '9':
     case ':':
@@ -715,12 +735,12 @@ next:
          * digits followed by a colon, but identifiers never do. */
         p = (p[strspn(p, "0123456789abcdefABCDEF")] == ':'
              ? lex_parse_integer(p, token)
-             : lex_parse_id(p, token));
+             : lex_parse_id(p, LEX_T_ID, token));
         break;
 
     default:
         if (lex_is_id1(*p)) {
-            p = lex_parse_id(p, token);
+            p = lex_parse_id(p, LEX_T_ID, token);
         } else {
             if (isprint((unsigned char) *p)) {
                 lex_error(token, "Invalid character `%c' in input.", *p);

ovn/lib/lex.h

@@ -36,6 +36,7 @@ enum lex_type {
     LEX_T_STRING,               /* "foo" */
     LEX_T_INTEGER,              /* 12345 or 1.2.3.4 or ::1 or 01:02:03:04:05 */
     LEX_T_MASKED_INTEGER,       /* 12345/10 or 1.2.0.0/16 or ::2/127 or... */
+    LEX_T_MACRO,                /* $NAME */
     LEX_T_ERROR,                /* invalid input */
 
     /* Bare tokens. */
@@ -77,20 +78,32 @@ enum lex_format {
 };
 const char *lex_format_to_string(enum lex_format);
 
-/* A token.
+/* A token. */
- *
- * 's' may point to 'buffer'; otherwise, it points to malloc()ed memory owned
- * by the token. */
 struct lex_token {
-    enum lex_type type;         /* One of LEX_*. */
+    /* One of LEX_*. */
-    char *s;                    /* LEX_T_ID, LEX_T_STRING, LEX_T_ERROR only. */
+    enum lex_type type;
-    enum lex_format format;     /* LEX_T_INTEGER, LEX_T_MASKED_INTEGER only. */
+
+    /* Meaningful for LEX_T_ID, LEX_T_STRING, LEX_T_ERROR, LEX_T_MACRO only.
+     * For these token types, 's' may point to 'buffer'; otherwise, it points
+     * to malloc()ed memory owned by the token.
+     *
+     * Must be NULL for other token types.
+     *
+     * For LEX_T_MACRO, 's' does not include the leading $. */
+    char *s;
+
+    /* LEX_T_INTEGER, LEX_T_MASKED_INTEGER only. */
+    enum lex_format format;
+
     union {
+        /* LEX_T_INTEGER, LEX_T_MASKED_INTEGER only. */
         struct {
             union mf_subvalue value; /* LEX_T_INTEGER, LEX_T_MASKED_INTEGER. */
             union mf_subvalue mask;  /* LEX_T_MASKED_INTEGER only. */
         };
-        char buffer[256];            /* Buffer for LEX_T_ID/LEX_T_STRING. */
+
+        /* LEX_T_ID, LEX_T_STRING, LEX_T_ERROR, LEX_T_MACRO only. */
+        char buffer[256];
     };
 };
 

tests/ovn.at

@@ -8,6 +8,9 @@ foo bar baz quuxquuxquux _abcd_ a.b.c.d a123_.456
 "abc\u0020def" => "abc def"
 " => error("Input ends inside quoted string.")dnl "
 
+$foo $bar $baz $quuxquuxquux $_abcd_ $a.b.c.d $a123_.456
+$1 => error("`$' must be followed by a valid identifier.") 1
+
 a/*b*/c => a c
 a//b c => a
 a/**/b => a b
@@ -215,6 +218,8 @@ eth.type != 0x800 => Nominal field eth.type may only be tested for equality (tak
 
 123 == 123 => Syntax error at `123' expecting field name.
 
+$name => Syntax error at `$name' expecting address set name.
+
 123 == xyzzy => Syntax error at `xyzzy' expecting field name.
 xyzzy == 1 => Syntax error at `xyzzy' expecting field name.
 
@@ -434,6 +439,56 @@ AT_CHECK([expr_to_flow 'inport == "eth0" && inport == "eth1"'], [0], [dnl
 ])
 AT_CLEANUP
 
+AT_SETUP([ovn -- converting expressions to flows -- address sets])
+expr_to_flow () {
+    echo "$1" | ovstest test-ovn expr-to-flows | sort
+}
+AT_CHECK([expr_to_flow 'ip4.src == {10.0.0.1, 10.0.0.2, 10.0.0.3}'], [0], [dnl
+ip,nw_src=10.0.0.1
+ip,nw_src=10.0.0.2
+ip,nw_src=10.0.0.3
+])
+AT_CHECK([expr_to_flow 'ip4.src == $set1'], [0], [dnl
+ip,nw_src=10.0.0.1
+ip,nw_src=10.0.0.2
+ip,nw_src=10.0.0.3
+])
+AT_CHECK([expr_to_flow 'ip4.src == {1.2.3.4, $set1}'], [0], [dnl
+ip,nw_src=1.2.3.4
+ip,nw_src=10.0.0.1
+ip,nw_src=10.0.0.2
+ip,nw_src=10.0.0.3
+])
+AT_CHECK([expr_to_flow 'ip4.src == {1.2.0.0/20, 5.5.5.0/24, $set1}'], [0], [dnl
+ip,nw_src=1.2.0.0/20
+ip,nw_src=10.0.0.1
+ip,nw_src=10.0.0.2
+ip,nw_src=10.0.0.3
+ip,nw_src=5.5.5.0/24
+])
+AT_CHECK([expr_to_flow 'ip6.src == {::1, ::2, ::3}'], [0], [dnl
+ipv6,ipv6_src=::1
+ipv6,ipv6_src=::2
+ipv6,ipv6_src=::3
+])
+AT_CHECK([expr_to_flow 'ip6.src == {::1, $set2, ::4}'], [0], [dnl
+ipv6,ipv6_src=::1
+ipv6,ipv6_src=::2
+ipv6,ipv6_src=::3
+ipv6,ipv6_src=::4
+])
+AT_CHECK([expr_to_flow 'eth.src == {00:00:00:00:00:01, 00:00:00:00:00:02, 00:00:00:00:00:03}'], [0], [dnl
+dl_src=00:00:00:00:00:01
+dl_src=00:00:00:00:00:02
+dl_src=00:00:00:00:00:03
+])
+AT_CHECK([expr_to_flow 'eth.src == {$set3}'], [0], [dnl
+dl_src=00:00:00:00:00:01
+dl_src=00:00:00:00:00:02
+dl_src=00:00:00:00:00:03
+])
+AT_CLEANUP
+
 AT_SETUP([ovn -- action parsing])
 dnl Text before => is input, text after => is expected output.
 AT_DATA([test-cases.txt], [[

tests/test-ovn.c

@@ -268,6 +268,26 @@ create_dhcp_opts(struct hmap *dhcp_opts)
     dhcp_opt_add(dhcp_opts, "lease_time",  51, "uint32");
 }
 
+static void
+create_macros(struct shash *macros)
+{
+    shash_init(macros);
+
+    static const char *const addrs1[] = {
+        "10.0.0.1", "10.0.0.2", "10.0.0.3",
+    };
+    static const char *const addrs2[] = {
+        "::1", "::2", "::3",
+    };
+    static const char *const addrs3[] = {
+        "00:00:00:00:00:01", "00:00:00:00:00:02", "00:00:00:00:00:03",
+    };
+
+    expr_macros_add(macros, "set1", addrs1, 3);
+    expr_macros_add(macros, "set2", addrs2, 3);
+    expr_macros_add(macros, "set3", addrs3, 3);
+}
+
 static bool
 lookup_port_cb(const void *ports_, const char *port_name, unsigned int *portp)
 {
@@ -284,10 +304,12 @@ static void
 test_parse_expr__(int steps)
 {
     struct shash symtab;
+    struct shash macros;
     struct simap ports;
     struct ds input;
 
     create_symtab(&symtab);
+    create_macros(&macros);
 
     simap_init(&ports);
     simap_put(&ports, "eth0", 5);
@@ -299,7 +321,7 @@ test_parse_expr__(int steps)
         struct expr *expr;
         char *error;
 
-        expr = expr_parse_string(ds_cstr(&input), &symtab, &error);
+        expr = expr_parse_string(ds_cstr(&input), &symtab, &macros, &error);
         if (!error && steps > 0) {
             expr = expr_annotate(expr, &symtab, &error);
         }
@@ -336,6 +358,8 @@ test_parse_expr__(int steps)
     simap_destroy(&ports);
     expr_symtab_destroy(&symtab);
     shash_destroy(&symtab);
+    expr_macros_destroy(&macros);
+    shash_destroy(&macros);
 }
 
 static void
@@ -480,7 +504,7 @@ test_evaluate_expr(struct ovs_cmdl_context *ctx)
         struct expr *expr;
         char *error;
 
-        expr = expr_parse_string(ds_cstr(&input), &symtab, &error);
+        expr = expr_parse_string(ds_cstr(&input), &symtab, NULL, &error);
         if (!error) {
             expr = expr_annotate(expr, &symtab, &error);
         }
@@ -954,7 +978,7 @@ test_tree_shape_exhaustively(struct expr *expr, struct shash *symtab,
             expr_format(expr, &s);
 
             char *error;
-            modified = expr_parse_string(ds_cstr(&s), symtab, &error);
+            modified = expr_parse_string(ds_cstr(&s), symtab, NULL, &error);
             if (error) {
                 fprintf(stderr, "%s fails to parse (%s)\n",
                         ds_cstr(&s), error);