implement new dns backend, res_send and other legacy resolver functions

this is the second phase of the "resolver overhaul" project.

the key additions in this commit are the __res_msend and __res_mkquery
functions, which have been factored so as to provide a backend for
both the legacy res_* functions and the standard getaddrinfo and
getnameinfo functions. the latter however are still using the old
backend code; there is code duplication which still needs to be
removed, and this will be the next phase of the resolver overhaul.

__res_msend is derived from the old __dns_doqueries function, but
generalized to send arbitrary caller-provided packets in parallel
rather than producing the parallel queries itself. this allows it to
be used (completely trivially) as a backend for res_send. the
factored-out query generation code, with slightly more generality, is
now part of __res_mkquery.

6 files changed, 263 insertions, 18 deletions

diff --git a/src/network/res_init.c b/src/network/res_init.c
index cbd5b155..5dba9dfc 100644
--- a/src/network/res_init.c
+++ b/src/network/res_init.c
@@ -1,3 +1,5 @@
+#include <resolv.h>
+
 int res_init()
 {
 	return 0;
diff --git a/src/network/res_mkquery.c b/src/network/res_mkquery.c
new file mode 100644
index 00000000..f7e4e9c6
--- /dev/null
+++ b/src/network/res_mkquery.c
@@ -0,0 +1,41 @@
+#include <resolv.h>
+#include <string.h>
+#include <time.h>
+#include "libc.h"
+
+int __res_mkquery(int op, const char *dname, int class, int type,
+	const unsigned char *data, int datalen,
+	const unsigned char *newrr, unsigned char *buf, int buflen)
+{
+	int id, i, j;
+	unsigned char q[280];
+	struct timespec ts;
+	size_t l = strnlen(dname, 256);
+
+	if (l-1>=254 || buflen<18+l || op>15u || class>255u || type>255u)
+		return -1;
+
+	/* Construct query template - ID will be filled later */
+	memset(q, 0, 18+l);
+	q[2] = op*8 + 1;
+	q[5] = 1;
+	memcpy((char *)q+13, dname, l);
+	for (i=13; q[i]; i=j+1) {
+		for (j=i; q[j] && q[j] != '.'; j++);
+		if (j-i-1u > 62u) return -1;
+		q[i-1] = j-i;
+	}
+	q[i+1] = type;
+	q[i+3] = class;
+
+	/* Make a reasonably unpredictable id */
+	clock_gettime(CLOCK_REALTIME, &ts);
+	id = ts.tv_nsec + ts.tv_nsec/65536UL & 0xffff;
+	q[0] = id/256;
+	q[1] = id;
+
+	memcpy(buf, q, 18+l);
+	return 18+l;
+}
+
+weak_alias(__res_mkquery, res_mkquery);
diff --git a/src/network/res_msend.c b/src/network/res_msend.c
new file mode 100644
index 00000000..cded5353
--- /dev/null
+++ b/src/network/res_msend.c
@@ -0,0 +1,184 @@
+#include <sys/socket.h>
+#include <netinet/in.h>
+#include <netdb.h>
+#include <arpa/inet.h>
+#include <stdint.h>
+#include <string.h>
+#include <poll.h>
+#include <time.h>
+#include <ctype.h>
+#include <unistd.h>
+#include <errno.h>
+#include <pthread.h>
+#include "stdio_impl.h"
+#include "syscall.h"
+
+static void cleanup(void *p)
+{
+	__syscall(SYS_close, (intptr_t)p);
+}
+
+static unsigned long mtime()
+{
+	struct timespec ts;
+	clock_gettime(CLOCK_REALTIME, &ts);
+	return (unsigned long)ts.tv_sec * 1000
+		+ ts.tv_nsec / 1000000;
+}
+
+int __res_msend(int nqueries, const unsigned char *const *queries,
+	const int *qlens, unsigned char *const *answers, int *alens, int asize)
+{
+	int fd;
+	FILE *f, _f;
+	unsigned char _buf[256];
+	char line[64], *s, *z;
+	int timeout = 5000, attempts = 2, retry_interval;
+	union {
+		struct sockaddr_in sin;
+		struct sockaddr_in6 sin6;
+	} sa = {0}, ns[3] = {{0}};
+	socklen_t sl = sizeof sa.sin;
+	int nns = 0;
+	int family = AF_INET;
+	int rlen;
+	int next;
+	int i, j;
+	int cs;
+	struct pollfd pfd;
+	unsigned long t0, t1, t2;
+
+	pthread_setcancelstate(PTHREAD_CANCEL_DISABLE, &cs);
+
+	/* Get nameservers from resolv.conf, fallback to localhost */
+	f = __fopen_rb_ca("/etc/resolv.conf", &_f, _buf, sizeof _buf);
+	if (f) for (nns=0; nns<3 && fgets(line, sizeof line, f); ) {
+		if (!strncmp(line, "options", 7) && isspace(line[7])) {
+			unsigned long x;
+			char *p, *z;
+			p = strstr(line, "timeout:");
+			if (p && isdigit(p[8])) {
+				p += 8;
+				x = strtoul(p, &z, 10);
+				if (z != p) timeout = x < 30 ? x*1000 : 30000;
+			}
+			p = strstr(line, "attempts:");
+			if (p && isdigit(p[9])) {
+				p += 9;
+				x = strtoul(p, &z, 10);
+				if (z != p) attempts = x < 10 ? x : 10;
+				if (!attempts) attempts = 1;
+			}
+		}
+		if (strncmp(line, "nameserver", 10) || !isspace(line[10]))
+			continue;
+		for (s=line+11; isspace(*s); s++);
+		for (z=s; *z && !isspace(*z); z++);
+		*z=0;
+
+		if (inet_pton(AF_INET, s, &ns[nns].sin.sin_addr)>0) {
+			ns[nns].sin.sin_port = htons(53);
+			ns[nns++].sin.sin_family = AF_INET;
+		} else if (inet_pton(AF_INET6, s, &ns[nns].sin6.sin6_addr)>0) {
+			sl = sizeof sa.sin6;
+			ns[nns].sin6.sin6_port = htons(53);
+			ns[nns++].sin6.sin6_family = family = AF_INET6;
+		}
+	}
+	if (f) __fclose_ca(f);
+	if (!nns) {
+		ns[0].sin.sin_family = AF_INET;
+		ns[0].sin.sin_port = htons(53);
+		ns[0].sin.sin_addr.s_addr = htonl(0x7f000001);
+		nns=1;
+	}
+
+	/* Get local address and open/bind a socket */
+	sa.sin.sin_family = family;
+	fd = socket(family, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0);
+
+	/* Handle case where system lacks IPv6 support */
+	if (fd < 0 && family == AF_INET6 && errno == EAFNOSUPPORT) {
+		fd = socket(AF_INET, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0);
+		family = AF_INET;
+	}
+	if (fd < 0 || bind(fd, (void *)&sa, sl) < 0) return -1;
+
+	/* Past this point, there are no errors. Each individual query will
+	 * yield either no reply (indicated by zero length) or an answer
+	 * packet which is up to the caller to interpret. */
+
+	pthread_cleanup_push(cleanup, (void *)(intptr_t)fd);
+	pthread_setcancelstate(cs, 0);
+
+	/* Convert any IPv4 addresses in a mixed environment to v4-mapped */
+	if (family == AF_INET6) {
+		setsockopt(fd, IPPROTO_IPV6, IPV6_V6ONLY, &(int){0}, sizeof 0);
+		for (i=0; i<nns; i++) {
+			if (ns[i].sin.sin_family != AF_INET) continue;
+			memcpy(ns[i].sin6.sin6_addr.s6_addr+12,
+				&ns[i].sin.sin_addr, 4);
+			memcpy(ns[i].sin6.sin6_addr.s6_addr,
+				"\0\0\0\0\0\0\0\0\0\0\xff\xff", 12);
+			ns[i].sin6.sin6_family = AF_INET6;
+			ns[i].sin6.sin6_flowinfo = 0;
+			ns[i].sin6.sin6_scope_id = 0;
+		}
+	}
+
+	memset(alens, 0, sizeof *alens * nqueries);
+
+	pfd.fd = fd;
+	pfd.events = POLLIN;
+	retry_interval = timeout / attempts;
+	t0 = t2 = mtime();
+	t1 = t2 - retry_interval;
+
+	for (; t2-t0 < timeout; t2=mtime()) {
+		if (t2-t1 >= retry_interval) {
+			/* Query all configured namservers in parallel */
+			for (i=0; i<nqueries; i++)
+				if (!alens[i])
+					for (j=0; j<nns; j++)
+						sendto(fd, queries[i],
+							qlens[i], MSG_NOSIGNAL,
+							(void *)&ns[j], sl);
+			t1 = t2;
+		}
+
+		/* Wait for a response, or until time to retry */
+		if (poll(&pfd, 1, t1+retry_interval-t2) <= 0) continue;
+
+		while ((rlen = recvfrom(fd, answers[next], asize, 0,
+		  (void *)&sa, (socklen_t[1]){sl})) >= 0) {
+
+			/* Ignore non-identifiable packets (no query id) */
+			if (rlen < 2) continue;
+
+			/* Ignore replies from addresses we didn't send to */
+			for (i=0; i<nns && memcmp(ns+i, &sa, sl); i++);
+			if (i==nns) continue;
+
+			/* Find which query this answer goes with, if any */
+			for (i=next; i<nqueries && (
+				answers[next][0] != queries[i][0] ||
+				answers[next][1] != queries[i][1] ); i++);
+			if (i==nqueries) continue;
+			if (alens[i]) continue;
+
+			/* Store answer in the right slot, or update next
+			 * available temp slot if it's already in place. */
+			alens[i] = rlen;
+			if (i == next)
+				for (; next<nqueries && alens[next]; next++);
+			else
+				memcpy(answers[i], answers[next], rlen);
+
+			if (next == nqueries) goto out;
+		}
+	}
+out:
+	pthread_cleanup_pop(1);
+
+	return 0;
+}
diff --git a/src/network/res_query.c b/src/network/res_query.c
index 3847da37..2b4e4bb1 100644
--- a/src/network/res_query.c
+++ b/src/network/res_query.c
@@ -1,25 +1,17 @@
-#define _GNU_SOURCE
 #include <resolv.h>
 #include <netdb.h>
-#include "__dns.h"
 #include "libc.h"
 
-int res_query(const char *name, int class, int type, unsigned char *dest, int len)
+int __res_mkquery(int, const char *, int, int, const unsigned char *, int, const unsigned char*, unsigned char *, int);
+int __res_send(const unsigned char *, int, unsigned char *, int);
+
+int __res_query(const char *name, int class, int type, unsigned char *dest, int len)
 {
-	if (class != 1 || len < 512)
-		return -1;
-	switch(__dns_doqueries(dest, name, &type, 1)) {
-	case EAI_NONAME:
-		h_errno = HOST_NOT_FOUND;
-		return -1;
-	case EAI_AGAIN:
-		h_errno = TRY_AGAIN;
-		return -1;
-	case EAI_FAIL:
-		h_errno = NO_RECOVERY;
-		return -1;
-	}
-	return 512;
+	unsigned char q[280];
+	int ql = __res_mkquery(0, name, class, type, 0, 0, 0, q, sizeof q);
+	if (ql < 0) return ql;
+	return __res_send(q, ql, dest, len);
 }
 
-weak_alias(res_query, res_search);
+weak_alias(__res_query, res_query);
+weak_alias(__res_query, res_search);
diff --git a/src/network/res_querydomain.c b/src/network/res_querydomain.c
new file mode 100644
index 00000000..c746dbe6
--- /dev/null
+++ b/src/network/res_querydomain.c
@@ -0,0 +1,14 @@
+#include <resolv.h>
+#include <string.h>
+
+int res_querydomain(const char *name, const char *domain, int class, int type, unsigned char *dest, int len)
+{
+	char tmp[256];
+	size_t nl = strnlen(name, 256);
+	size_t dl = strnlen(domain, 256);
+	if (nl+dl+1 > 255) return -1;
+	memcpy(tmp, name, nl);
+	tmp[nl] = '.';
+	memcpy(tmp+nl+1, domain, dl+1);
+	return res_query(tmp, class, type, dest, len);
+}
diff --git a/src/network/res_send.c b/src/network/res_send.c
new file mode 100644
index 00000000..19cfe0f6
--- /dev/null
+++ b/src/network/res_send.c
@@ -0,0 +1,12 @@
+#include <resolv.h>
+#include "libc.h"
+
+int __res_msend(int, const unsigned char *const *, const int *, unsigned char *const *, int *, int);
+
+int __res_send(const unsigned char *msg, int msglen, unsigned char *answer, int anslen)
+{
+	int r = __res_msend(1, &msg, &msglen, &answer, &anslen, anslen);
+	return r<0 ? r : anslen;
+}
+
+weak_alias(__res_send, res_send);