diff options
Diffstat (limited to 'src/network')
29 files changed, 376 insertions, 95 deletions
diff --git a/src/network/accept4.c b/src/network/accept4.c index 59ab1726..765a38ed 100644 --- a/src/network/accept4.c +++ b/src/network/accept4.c @@ -9,6 +9,10 @@ int accept4(int fd, struct sockaddr *restrict addr, socklen_t *restrict len, int if (!flg) return accept(fd, addr, len); int ret = socketcall_cp(accept4, fd, addr, len, flg, 0, 0); if (ret>=0 || (errno != ENOSYS && errno != EINVAL)) return ret; + if (flg & ~(SOCK_CLOEXEC|SOCK_NONBLOCK)) { + errno = EINVAL; + return -1; + } ret = accept(fd, addr, len); if (ret<0) return ret; if (flg & SOCK_CLOEXEC) diff --git a/src/network/dns_parse.c b/src/network/dns_parse.c index e6ee19d9..09813112 100644 --- a/src/network/dns_parse.c +++ b/src/network/dns_parse.c @@ -1,7 +1,7 @@ #include <string.h> #include "lookup.h" -int __dns_parse(const unsigned char *r, int rlen, int (*callback)(void *, int, const void *, int, const void *), void *ctx) +int __dns_parse(const unsigned char *r, int rlen, int (*callback)(void *, int, const void *, int, const void *, int), void *ctx) { int qdcount, ancount; const unsigned char *p; @@ -12,21 +12,20 @@ int __dns_parse(const unsigned char *r, int rlen, int (*callback)(void *, int, c p = r+12; qdcount = r[4]*256 + r[5]; ancount = r[6]*256 + r[7]; - if (qdcount+ancount > 64) return -1; while (qdcount--) { while (p-r < rlen && *p-1U < 127) p++; - if (*p>193 || (*p==193 && p[1]>254) || p>r+rlen-6) + if (p>r+rlen-6) return -1; p += 5 + !!*p; } while (ancount--) { while (p-r < rlen && *p-1U < 127) p++; - if (*p>193 || (*p==193 && p[1]>254) || p>r+rlen-6) + if (p>r+rlen-12) return -1; p += 1 + !!*p; len = p[8]*256 + p[9]; - if (p+len > r+rlen) return -1; - if (callback(ctx, p[1], p+10, len, r) < 0) return -1; + if (len+10 > r+rlen-p) return -1; + if (callback(ctx, p[1], p+10, len, r, rlen) < 0) return -1; p += 10 + len; } return 0; diff --git a/src/network/gai_strerror.c b/src/network/gai_strerror.c index 9596580e..56b71503 100644 --- a/src/network/gai_strerror.c +++ b/src/network/gai_strerror.c @@ -6,7 +6,7 @@ static const char msgs[] = "Name does not resolve\0" "Try again\0" "Non-recoverable error\0" - "Unknown error\0" + "Name has no usable address\0" "Unrecognized address family or invalid length\0" "Unrecognized socket type\0" "Unrecognized service\0" diff --git a/src/network/getaddrinfo.c b/src/network/getaddrinfo.c index efaab306..64ad259a 100644 --- a/src/network/getaddrinfo.c +++ b/src/network/getaddrinfo.c @@ -16,6 +16,7 @@ int getaddrinfo(const char *restrict host, const char *restrict serv, const stru char canon[256], *outcanon; int nservs, naddrs, nais, canon_len, i, j, k; int family = AF_UNSPEC, flags = 0, proto = 0, socktype = 0; + int no_family = 0; struct aibuf *out; if (!host && !serv) return EAI_NONAME; @@ -66,9 +67,11 @@ int getaddrinfo(const char *restrict host, const char *restrict serv, const stru pthread_setcancelstate( PTHREAD_CANCEL_DISABLE, &cs); int r = connect(s, ta[i], tl[i]); + int saved_errno = errno; pthread_setcancelstate(cs, 0); close(s); if (!r) continue; + errno = saved_errno; } switch (errno) { case EADDRNOTAVAIL: @@ -80,7 +83,7 @@ int getaddrinfo(const char *restrict host, const char *restrict serv, const stru default: return EAI_SYSTEM; } - if (family == tf[i]) return EAI_NONAME; + if (family == tf[i]) no_family = 1; family = tf[1-i]; } } @@ -91,6 +94,8 @@ int getaddrinfo(const char *restrict host, const char *restrict serv, const stru naddrs = __lookup_name(addrs, canon, host, family, flags); if (naddrs < 0) return naddrs; + if (no_family) return EAI_NODATA; + nais = nservs * naddrs; canon_len = strlen(canon); out = calloc(1, nais * sizeof(*out) + canon_len + 1); diff --git a/src/network/gethostbyaddr.c b/src/network/gethostbyaddr.c index 598e2241..c3cacaac 100644 --- a/src/network/gethostbyaddr.c +++ b/src/network/gethostbyaddr.c @@ -20,5 +20,5 @@ struct hostent *gethostbyaddr(const void *a, socklen_t l, int af) err = gethostbyaddr_r(a, l, af, h, (void *)(h+1), size-sizeof *h, &res, &h_errno); } while (err == ERANGE); - return err ? 0 : h; + return res; } diff --git a/src/network/gethostbyaddr_r.c b/src/network/gethostbyaddr_r.c index 0f1e61aa..ceaf3935 100644 --- a/src/network/gethostbyaddr_r.c +++ b/src/network/gethostbyaddr_r.c @@ -54,10 +54,11 @@ int gethostbyaddr_r(const void *a, socklen_t l, int af, case EAI_OVERFLOW: return ERANGE; default: - case EAI_MEMORY: - case EAI_SYSTEM: case EAI_FAIL: *err = NO_RECOVERY; + return EBADMSG; + case EAI_SYSTEM: + *err = NO_RECOVERY; return errno; case 0: break; diff --git a/src/network/gethostbyname2.c b/src/network/gethostbyname2.c index dc9d6621..bd0da7f8 100644 --- a/src/network/gethostbyname2.c +++ b/src/network/gethostbyname2.c @@ -21,5 +21,5 @@ struct hostent *gethostbyname2(const char *name, int af) err = gethostbyname2_r(name, af, h, (void *)(h+1), size-sizeof *h, &res, &h_errno); } while (err == ERANGE); - return err ? 0 : h; + return res; } diff --git a/src/network/gethostbyname2_r.c b/src/network/gethostbyname2_r.c index fc894877..a5eb67fe 100644 --- a/src/network/gethostbyname2_r.c +++ b/src/network/gethostbyname2_r.c @@ -22,7 +22,10 @@ int gethostbyname2_r(const char *name, int af, if (cnt<0) switch (cnt) { case EAI_NONAME: *err = HOST_NOT_FOUND; - return ENOENT; + return 0; + case EAI_NODATA: + *err = NO_DATA; + return 0; case EAI_AGAIN: *err = TRY_AGAIN; return EAGAIN; @@ -30,7 +33,6 @@ int gethostbyname2_r(const char *name, int af, case EAI_FAIL: *err = NO_RECOVERY; return EBADMSG; - case EAI_MEMORY: case EAI_SYSTEM: *err = NO_RECOVERY; return errno; diff --git a/src/network/getifaddrs.c b/src/network/getifaddrs.c index fed75bd8..74df4d6c 100644 --- a/src/network/getifaddrs.c +++ b/src/network/getifaddrs.c @@ -39,8 +39,8 @@ struct ifaddrs_storage { }; struct ifaddrs_ctx { - struct ifaddrs_storage *first; - struct ifaddrs_storage *last; + struct ifaddrs *first; + struct ifaddrs *last; struct ifaddrs_storage *hash[IFADDRS_HASH_SIZE]; }; @@ -195,9 +195,9 @@ static int netlink_msg_to_ifaddr(void *pctx, struct nlmsghdr *h) } if (ifs->ifa.ifa_name) { - if (!ctx->first) ctx->first = ifs; - if (ctx->last) ctx->last->ifa.ifa_next = &ifs->ifa; - ctx->last = ifs; + if (!ctx->first) ctx->first = &ifs->ifa; + if (ctx->last) ctx->last->ifa_next = &ifs->ifa; + ctx->last = &ifs->ifa; } else { free(ifs); } @@ -210,7 +210,7 @@ int getifaddrs(struct ifaddrs **ifap) int r; memset(ctx, 0, sizeof *ctx); r = __rtnetlink_enumerate(AF_UNSPEC, AF_UNSPEC, netlink_msg_to_ifaddr, ctx); - if (r == 0) *ifap = &ctx->first->ifa; - else freeifaddrs(&ctx->first->ifa); + if (r == 0) *ifap = ctx->first; + else freeifaddrs(ctx->first); return r; } diff --git a/src/network/getnameinfo.c b/src/network/getnameinfo.c index f77e73ad..133c15b3 100644 --- a/src/network/getnameinfo.c +++ b/src/network/getnameinfo.c @@ -58,6 +58,7 @@ static void reverse_hosts(char *buf, const unsigned char *a, unsigned scopeid, i if ((p=strchr(line, '#'))) *p++='\n', *p=0; for (p=line; *p && !isspace(*p); p++); + if (!*p) continue; *p++ = 0; if (__lookup_ipliteral(&iplit, line, AF_UNSPEC)<=0) continue; @@ -108,10 +109,10 @@ static void reverse_services(char *buf, int port, int dgram) __fclose_ca(f); } -static int dns_parse_callback(void *c, int rr, const void *data, int len, const void *packet) +static int dns_parse_callback(void *c, int rr, const void *data, int len, const void *packet, int plen) { if (rr != RR_PTR) return 0; - if (__dn_expand(packet, (const unsigned char *)packet + 512, + if (__dn_expand(packet, (const unsigned char *)packet + plen, data, c, 256) <= 0) *(char *)c = 0; return 0; @@ -158,10 +159,13 @@ int getnameinfo(const struct sockaddr *restrict sa, socklen_t sl, unsigned char query[18+PTR_MAX], reply[512]; int qlen = __res_mkquery(0, ptr, 1, RR_PTR, 0, 0, 0, query, sizeof query); + query[3] = 0; /* don't need AD flag */ int rlen = __res_send(query, qlen, reply, sizeof reply); buf[0] = 0; - if (rlen > 0) + if (rlen > 0) { + if (rlen > sizeof reply) rlen = sizeof reply; __dns_parse(reply, rlen, dns_parse_callback, buf); + } } if (!*buf) { if (flags & NI_NAMEREQD) return EAI_NONAME; diff --git a/src/network/getservbyport_r.c b/src/network/getservbyport_r.c index b7f21c6b..e4cc3079 100644 --- a/src/network/getservbyport_r.c +++ b/src/network/getservbyport_r.c @@ -26,7 +26,7 @@ int getservbyport_r(int port, const char *prots, /* Align buffer */ i = (uintptr_t)buf & sizeof(char *)-1; if (!i) i = sizeof(char *); - if (buflen < 3*sizeof(char *)-i) + if (buflen <= 3*sizeof(char *)-i) return ERANGE; buf += sizeof(char *)-i; buflen -= sizeof(char *)-i; @@ -46,6 +46,8 @@ int getservbyport_r(int port, const char *prots, case EAI_MEMORY: case EAI_SYSTEM: return ENOMEM; + case EAI_OVERFLOW: + return ERANGE; default: return ENOENT; case 0: diff --git a/src/network/getsockopt.c b/src/network/getsockopt.c index e871d624..d3640d9c 100644 --- a/src/network/getsockopt.c +++ b/src/network/getsockopt.c @@ -26,6 +26,15 @@ int getsockopt(int fd, int level, int optname, void *restrict optval, socklen_t tv->tv_sec = tv32[0]; tv->tv_usec = tv32[1]; *optlen = sizeof *tv; + break; + case SO_TIMESTAMP: + case SO_TIMESTAMPNS: + if (SO_TIMESTAMP == SO_TIMESTAMP_OLD) break; + if (optname==SO_TIMESTAMP) optname=SO_TIMESTAMP_OLD; + if (optname==SO_TIMESTAMPNS) optname=SO_TIMESTAMPNS_OLD; + r = __socketcall(getsockopt, fd, level, + optname, optval, optlen, 0); + break; } } return __syscall_ret(r); diff --git a/src/network/h_errno.c b/src/network/h_errno.c index 4f700cea..638f7718 100644 --- a/src/network/h_errno.c +++ b/src/network/h_errno.c @@ -1,9 +1,11 @@ #include <netdb.h> +#include "pthread_impl.h" #undef h_errno int h_errno; int *__h_errno_location(void) { - return &h_errno; + if (!__pthread_self()->stack) return &h_errno; + return &__pthread_self()->h_errno_val; } diff --git a/src/network/herror.c b/src/network/herror.c index 65f25ff3..87f8cff4 100644 --- a/src/network/herror.c +++ b/src/network/herror.c @@ -4,5 +4,5 @@ void herror(const char *msg) { - fprintf(stderr, "%s%s%s", msg?msg:"", msg?": ":"", hstrerror(h_errno)); + fprintf(stderr, "%s%s%s\n", msg?msg:"", msg?": ":"", hstrerror(h_errno)); } diff --git a/src/network/inet_ntop.c b/src/network/inet_ntop.c index 4bfef2c5..f442f47d 100644 --- a/src/network/inet_ntop.c +++ b/src/network/inet_ntop.c @@ -34,7 +34,12 @@ const char *inet_ntop(int af, const void *restrict a0, char *restrict s, socklen for (i=best=0, max=2; buf[i]; i++) { if (i && buf[i] != ':') continue; j = strspn(buf+i, ":0"); - if (j>max) best=i, max=j; + /* The leading sequence of zeros (best==0) is + * disadvantaged compared to sequences elsewhere + * as it doesn't have a leading colon. One extra + * character is required for another sequence to + * beat it fairly. */ + if (j>max+(best==0)) best=i, max=j; } if (max>3) { buf[best] = buf[best+1] = ':'; diff --git a/src/network/inet_pton.c b/src/network/inet_pton.c index d36c3689..bcbdd9ef 100644 --- a/src/network/inet_pton.c +++ b/src/network/inet_pton.c @@ -54,6 +54,7 @@ int inet_pton(int af, const char *restrict s, void *restrict a0) if (s[j]!='.' || (i<6 && brk<0)) return 0; need_v4=1; i++; + ip[i&7]=0; break; } s += j+1; diff --git a/src/network/lookup.h b/src/network/lookup.h index ef662725..54b2f8b5 100644 --- a/src/network/lookup.h +++ b/src/network/lookup.h @@ -50,6 +50,6 @@ hidden int __lookup_ipliteral(struct address buf[static 1], const char *name, in hidden int __get_resolv_conf(struct resolvconf *, char *, size_t); hidden int __res_msend_rc(int, const unsigned char *const *, const int *, unsigned char *const *, int *, int, const struct resolvconf *); -hidden int __dns_parse(const unsigned char *, int, int (*)(void *, int, const void *, int, const void *), void *); +hidden int __dns_parse(const unsigned char *, int, int (*)(void *, int, const void *, int, const void *, int), void *); #endif diff --git a/src/network/lookup_ipliteral.c b/src/network/lookup_ipliteral.c index 2fddab73..1e766206 100644 --- a/src/network/lookup_ipliteral.c +++ b/src/network/lookup_ipliteral.c @@ -15,7 +15,7 @@ int __lookup_ipliteral(struct address buf[static 1], const char *name, int famil struct in6_addr a6; if (__inet_aton(name, &a4) > 0) { if (family == AF_INET6) /* wrong family */ - return EAI_NONAME; + return EAI_NODATA; memcpy(&buf[0].addr, &a4, sizeof a4); buf[0].family = AF_INET; buf[0].scopeid = 0; @@ -34,7 +34,7 @@ int __lookup_ipliteral(struct address buf[static 1], const char *name, int famil if (inet_pton(AF_INET6, name, &a6) <= 0) return 0; if (family == AF_INET) /* wrong family */ - return EAI_NONAME; + return EAI_NODATA; memcpy(&buf[0].addr, &a6, sizeof a6); buf[0].family = AF_INET6; diff --git a/src/network/lookup_name.c b/src/network/lookup_name.c index c93263a9..35218185 100644 --- a/src/network/lookup_name.c +++ b/src/network/lookup_name.c @@ -50,7 +50,7 @@ static int name_from_hosts(struct address buf[static MAXADDRS], char canon[stati { char line[512]; size_t l = strlen(name); - int cnt = 0, badfam = 0; + int cnt = 0, badfam = 0, have_canon = 0; unsigned char _buf[1032]; FILE _f, *f = __fopen_rb_ca("/etc/hosts", &_f, _buf, sizeof _buf); if (!f) switch (errno) { @@ -79,15 +79,20 @@ static int name_from_hosts(struct address buf[static MAXADDRS], char canon[stati case 0: continue; default: - badfam = EAI_NONAME; - continue; + badfam = EAI_NODATA; + break; } + if (have_canon) continue; + /* Extract first name as canonical name */ for (; *p && isspace(*p); p++); for (z=p; *z && !isspace(*z); z++); *z = 0; - if (is_valid_hostname(p)) memcpy(canon, p, z-p+1); + if (is_valid_hostname(p)) { + have_canon = 1; + memcpy(canon, p, z-p+1); + } } __fclose_ca(f); return cnt ? cnt : badfam; @@ -97,45 +102,50 @@ struct dpc_ctx { struct address *addrs; char *canon; int cnt; + int rrtype; }; #define RR_A 1 #define RR_CNAME 5 #define RR_AAAA 28 -static int dns_parse_callback(void *c, int rr, const void *data, int len, const void *packet) +#define ABUF_SIZE 4800 + +static int dns_parse_callback(void *c, int rr, const void *data, int len, const void *packet, int plen) { char tmp[256]; + int family; struct dpc_ctx *ctx = c; - if (ctx->cnt >= MAXADDRS) return -1; + if (rr == RR_CNAME) { + if (__dn_expand(packet, (const unsigned char *)packet + plen, + data, tmp, sizeof tmp) > 0 && is_valid_hostname(tmp)) + strcpy(ctx->canon, tmp); + return 0; + } + if (ctx->cnt >= MAXADDRS) return 0; + if (rr != ctx->rrtype) return 0; switch (rr) { case RR_A: if (len != 4) return -1; - ctx->addrs[ctx->cnt].family = AF_INET; - ctx->addrs[ctx->cnt].scopeid = 0; - memcpy(ctx->addrs[ctx->cnt++].addr, data, 4); + family = AF_INET; break; case RR_AAAA: if (len != 16) return -1; - ctx->addrs[ctx->cnt].family = AF_INET6; - ctx->addrs[ctx->cnt].scopeid = 0; - memcpy(ctx->addrs[ctx->cnt++].addr, data, 16); - break; - case RR_CNAME: - if (__dn_expand(packet, (const unsigned char *)packet + 512, - data, tmp, sizeof tmp) > 0 && is_valid_hostname(tmp)) - strcpy(ctx->canon, tmp); + family = AF_INET6; break; } + ctx->addrs[ctx->cnt].family = family; + ctx->addrs[ctx->cnt].scopeid = 0; + memcpy(ctx->addrs[ctx->cnt++].addr, data, len); return 0; } static int name_from_dns(struct address buf[static MAXADDRS], char canon[static 256], const char *name, int family, const struct resolvconf *conf) { - unsigned char qbuf[2][280], abuf[2][512]; + unsigned char qbuf[2][280], abuf[2][ABUF_SIZE]; const unsigned char *qp[2] = { qbuf[0], qbuf[1] }; unsigned char *ap[2] = { abuf[0], abuf[1] }; - int qlens[2], alens[2]; + int qlens[2], alens[2], qtypes[2]; int i, nq = 0; struct dpc_ctx ctx = { .addrs = buf, .canon = canon }; static const struct { int af; int rr; } afrr[2] = { @@ -148,7 +158,12 @@ static int name_from_dns(struct address buf[static MAXADDRS], char canon[static qlens[nq] = __res_mkquery(0, name, 1, afrr[i].rr, 0, 0, 0, qbuf[nq], sizeof *qbuf); if (qlens[nq] == -1) - return EAI_NONAME; + return 0; + qtypes[nq] = afrr[i].rr; + qbuf[nq][3] = 0; /* don't need AD flag */ + /* Ensure query IDs are distinct. */ + if (nq && qbuf[nq][0] == qbuf[0][0]) + qbuf[nq][0]++; nq++; } } @@ -156,14 +171,20 @@ static int name_from_dns(struct address buf[static MAXADDRS], char canon[static if (__res_msend_rc(nq, qp, qlens, ap, alens, sizeof *abuf, conf) < 0) return EAI_SYSTEM; - for (i=0; i<nq; i++) + for (i=0; i<nq; i++) { + if (alens[i] < 4 || (abuf[i][3] & 15) == 2) return EAI_AGAIN; + if ((abuf[i][3] & 15) == 3) return 0; + if ((abuf[i][3] & 15) != 0) return EAI_FAIL; + } + + for (i=nq-1; i>=0; i--) { + ctx.rrtype = qtypes[i]; + if (alens[i] > sizeof(abuf[i])) alens[i] = sizeof abuf[i]; __dns_parse(abuf[i], alens[i], dns_parse_callback, &ctx); + } if (ctx.cnt) return ctx.cnt; - if (alens[0] < 4 || (abuf[0][3] & 15) == 2) return EAI_AGAIN; - if ((abuf[0][3] & 15) == 0) return EAI_NONAME; - if ((abuf[0][3] & 15) == 3) return 0; - return EAI_FAIL; + return EAI_NODATA; } static int name_from_dns_search(struct address buf[static MAXADDRS], char canon[static 256], const char *name, int family) diff --git a/src/network/netlink.h b/src/network/netlink.h index 38acb178..873fabe2 100644 --- a/src/network/netlink.h +++ b/src/network/netlink.h @@ -86,7 +86,7 @@ struct ifaddrmsg { #define RTA_DATALEN(rta) ((rta)->rta_len-sizeof(struct rtattr)) #define RTA_DATAEND(rta) ((char*)(rta)+(rta)->rta_len) #define RTA_NEXT(rta) (struct rtattr*)((char*)(rta)+NETLINK_ALIGN((rta)->rta_len)) -#define RTA_OK(nlh,end) ((char*)(end)-(char*)(rta) >= sizeof(struct rtattr)) +#define RTA_OK(rta,end) ((char*)(end)-(char*)(rta) >= sizeof(struct rtattr)) #define NLMSG_RTA(nlh,len) ((void*)((char*)(nlh)+sizeof(struct nlmsghdr)+NETLINK_ALIGN(len))) #define NLMSG_RTAOK(rta,nlh) RTA_OK(rta,NLMSG_DATAEND(nlh)) diff --git a/src/network/recvmmsg.c b/src/network/recvmmsg.c index d5dc6b51..2978e2f6 100644 --- a/src/network/recvmmsg.c +++ b/src/network/recvmmsg.c @@ -8,6 +8,8 @@ #define IS32BIT(x) !((x)+0x80000000ULL>>32) #define CLAMP(x) (int)(IS32BIT(x) ? (x) : 0x7fffffffU+((0ULL+(x))>>63)) +hidden void __convert_scm_timestamps(struct msghdr *, socklen_t); + int recvmmsg(int fd, struct mmsghdr *msgvec, unsigned int vlen, unsigned int flags, struct timespec *timeout) { #if LONG_MAX > INT_MAX @@ -19,14 +21,18 @@ int recvmmsg(int fd, struct mmsghdr *msgvec, unsigned int vlen, unsigned int fla #ifdef SYS_recvmmsg_time64 time_t s = timeout ? timeout->tv_sec : 0; long ns = timeout ? timeout->tv_nsec : 0; - int r = -ENOSYS; - if (SYS_recvmmsg == SYS_recvmmsg_time64 || !IS32BIT(s)) - r = __syscall_cp(SYS_recvmmsg_time64, fd, msgvec, vlen, flags, + int r = __syscall_cp(SYS_recvmmsg_time64, fd, msgvec, vlen, flags, timeout ? ((long long[]){s, ns}) : 0); if (SYS_recvmmsg == SYS_recvmmsg_time64 || r!=-ENOSYS) return __syscall_ret(r); - return syscall_cp(SYS_recvmmsg, fd, msgvec, vlen, flags, + if (vlen > IOV_MAX) vlen = IOV_MAX; + socklen_t csize[vlen]; + for (int i=0; i<vlen; i++) csize[i] = msgvec[i].msg_hdr.msg_controllen; + r = __syscall_cp(SYS_recvmmsg, fd, msgvec, vlen, flags, timeout ? ((long[]){CLAMP(s), ns}) : 0); + for (int i=0; i<r; i++) + __convert_scm_timestamps(&msgvec[i].msg_hdr, csize[i]); + return __syscall_ret(r); #else return syscall_cp(SYS_recvmmsg, fd, msgvec, vlen, flags, timeout); #endif diff --git a/src/network/recvmsg.c b/src/network/recvmsg.c index 4ca7da8b..03641625 100644 --- a/src/network/recvmsg.c +++ b/src/network/recvmsg.c @@ -1,10 +1,56 @@ #include <sys/socket.h> #include <limits.h> +#include <time.h> +#include <sys/time.h> +#include <string.h> #include "syscall.h" +hidden void __convert_scm_timestamps(struct msghdr *, socklen_t); + +void __convert_scm_timestamps(struct msghdr *msg, socklen_t csize) +{ + if (SCM_TIMESTAMP == SCM_TIMESTAMP_OLD) return; + if (!msg->msg_control || !msg->msg_controllen) return; + + struct cmsghdr *cmsg, *last=0; + long tmp; + long long tvts[2]; + int type = 0; + + for (cmsg=CMSG_FIRSTHDR(msg); cmsg; cmsg=CMSG_NXTHDR(msg, cmsg)) { + if (cmsg->cmsg_level==SOL_SOCKET) switch (cmsg->cmsg_type) { + case SCM_TIMESTAMP_OLD: + if (type) break; + type = SCM_TIMESTAMP; + goto common; + case SCM_TIMESTAMPNS_OLD: + type = SCM_TIMESTAMPNS; + common: + memcpy(&tmp, CMSG_DATA(cmsg), sizeof tmp); + tvts[0] = tmp; + memcpy(&tmp, CMSG_DATA(cmsg) + sizeof tmp, sizeof tmp); + tvts[1] = tmp; + break; + } + last = cmsg; + } + if (!last || !type) return; + if (CMSG_SPACE(sizeof tvts) > csize-msg->msg_controllen) { + msg->msg_flags |= MSG_CTRUNC; + return; + } + msg->msg_controllen += CMSG_SPACE(sizeof tvts); + cmsg = CMSG_NXTHDR(msg, last); + cmsg->cmsg_level = SOL_SOCKET; + cmsg->cmsg_type = type; + cmsg->cmsg_len = CMSG_LEN(sizeof tvts); + memcpy(CMSG_DATA(cmsg), &tvts, sizeof tvts); +} + ssize_t recvmsg(int fd, struct msghdr *msg, int flags) { ssize_t r; + socklen_t orig_controllen = msg->msg_controllen; #if LONG_MAX > INT_MAX struct msghdr h, *orig = msg; if (msg) { @@ -14,6 +60,7 @@ ssize_t recvmsg(int fd, struct msghdr *msg, int flags) } #endif r = socketcall_cp(recvmsg, fd, msg, flags, 0, 0, 0); + if (r >= 0) __convert_scm_timestamps(msg, orig_controllen); #if LONG_MAX > INT_MAX if (orig) *orig = h; #endif diff --git a/src/network/res_mkquery.c b/src/network/res_mkquery.c index 6fa04a5c..614bf786 100644 --- a/src/network/res_mkquery.c +++ b/src/network/res_mkquery.c @@ -13,6 +13,7 @@ int __res_mkquery(int op, const char *dname, int class, int type, int n; if (l && dname[l-1]=='.') l--; + if (l && dname[l-1]=='.') return -1; n = 17+l+!!l; if (l>253 || buflen<n || op>15u || class>255u || type>255u) return -1; @@ -20,6 +21,7 @@ int __res_mkquery(int op, const char *dname, int class, int type, /* Construct query template - ID will be filled later */ memset(q, 0, n); q[2] = op*8 + 1; + q[3] = 32; /* AD */ q[5] = 1; memcpy((char *)q+13, dname, l); for (i=13; q[i]; i=j+1) { diff --git a/src/network/res_msend.c b/src/network/res_msend.c index 3e018009..fcb52513 100644 --- a/src/network/res_msend.c +++ b/src/network/res_msend.c @@ -1,5 +1,6 @@ #include <sys/socket.h> #include <netinet/in.h> +#include <netinet/tcp.h> #include <netdb.h> #include <arpa/inet.h> #include <stdint.h> @@ -16,17 +17,65 @@ static void cleanup(void *p) { - __syscall(SYS_close, (intptr_t)p); + struct pollfd *pfd = p; + for (int i=0; pfd[i].fd >= -1; i++) + if (pfd[i].fd >= 0) __syscall(SYS_close, pfd[i].fd); } static unsigned long mtime() { struct timespec ts; - clock_gettime(CLOCK_REALTIME, &ts); + if (clock_gettime(CLOCK_MONOTONIC, &ts) < 0 && errno == ENOSYS) + clock_gettime(CLOCK_REALTIME, &ts); return (unsigned long)ts.tv_sec * 1000 + ts.tv_nsec / 1000000; } +static int start_tcp(struct pollfd *pfd, int family, const void *sa, socklen_t sl, const unsigned char *q, int ql) +{ + struct msghdr mh = { + .msg_name = (void *)sa, + .msg_namelen = sl, + .msg_iovlen = 2, + .msg_iov = (struct iovec [2]){ + { .iov_base = (uint8_t[]){ ql>>8, ql }, .iov_len = 2 }, + { .iov_base = (void *)q, .iov_len = ql } } + }; + int r; + int fd = socket(family, SOCK_STREAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); + pfd->fd = fd; + pfd->events = POLLOUT; + if (!setsockopt(fd, IPPROTO_TCP, TCP_FASTOPEN_CONNECT, + &(int){1}, sizeof(int))) { + r = sendmsg(fd, &mh, MSG_FASTOPEN|MSG_NOSIGNAL); + if (r == ql+2) pfd->events = POLLIN; + if (r >= 0) return r; + if (errno == EINPROGRESS) return 0; + } + r = connect(fd, sa, sl); + if (!r || errno == EINPROGRESS) return 0; + close(fd); + pfd->fd = -1; + return -1; +} + +static void step_mh(struct msghdr *mh, size_t n) +{ + /* Adjust iovec in msghdr to skip first n bytes. */ + while (mh->msg_iovlen && n >= mh->msg_iov->iov_len) { + n -= mh->msg_iov->iov_len; + mh->msg_iov++; + mh->msg_iovlen--; + } + if (!mh->msg_iovlen) return; + mh->msg_iov->iov_base = (char *)mh->msg_iov->iov_base + n; + mh->msg_iov->iov_len -= n; +} + +/* Internal contract for __res_msend[_rc]: asize must be >=512, nqueries + * must be sufficiently small to be safe as VLA size. In practice it's + * either 1 or 2, anyway. */ + int __res_msend_rc(int nqueries, const unsigned char *const *queries, const int *qlens, unsigned char *const *answers, int *alens, int asize, const struct resolvconf *conf) @@ -34,8 +83,8 @@ int __res_msend_rc(int nqueries, const unsigned char *const *queries, int fd; int timeout, attempts, retry_interval, servfail_retry; union { - struct sockaddr_in sin; struct sockaddr_in6 sin6; + struct sockaddr_in sin; } sa = {0}, ns[MAXNS] = {{0}}; socklen_t sl = sizeof sa.sin; int nns = 0; @@ -44,7 +93,10 @@ int __res_msend_rc(int nqueries, const unsigned char *const *queries, int next; int i, j; int cs; - struct pollfd pfd; + struct pollfd pfd[nqueries+2]; + int qpos[nqueries], apos[nqueries]; + unsigned char alen_buf[nqueries][2]; + int r; unsigned long t0, t1, t2; pthread_setcancelstate(PTHREAD_CANCEL_DISABLE, &cs); @@ -68,29 +120,22 @@ int __res_msend_rc(int nqueries, const unsigned char *const *queries, } /* 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) { + for (i=0; i<nns && conf->ns[nns].family == AF_INET6; i++); + if (i==nns) { + pthread_setcancelstate(cs, 0); + return -1; + } fd = socket(AF_INET, SOCK_DGRAM|SOCK_CLOEXEC|SOCK_NONBLOCK, 0); family = AF_INET; + sl = sizeof sa.sin; } - if (fd < 0 || bind(fd, (void *)&sa, sl) < 0) { - if (fd >= 0) close(fd); - pthread_setcancelstate(cs, 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) { + if (fd >= 0 && 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; @@ -104,16 +149,38 @@ int __res_msend_rc(int nqueries, const unsigned char *const *queries, } } + sa.sin.sin_family = family; + if (fd < 0 || bind(fd, (void *)&sa, sl) < 0) { + if (fd >= 0) close(fd); + pthread_setcancelstate(cs, 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. */ + + for (i=0; i<nqueries; i++) pfd[i].fd = -1; + pfd[nqueries].fd = fd; + pfd[nqueries].events = POLLIN; + pfd[nqueries+1].fd = -2; + + pthread_cleanup_push(cleanup, pfd); + pthread_setcancelstate(cs, 0); + memset(alens, 0, sizeof *alens * nqueries); - pfd.fd = fd; - pfd.events = POLLIN; retry_interval = timeout / attempts; next = 0; t0 = t2 = mtime(); t1 = t2 - retry_interval; for (; t2-t0 < timeout; t2=mtime()) { + /* This is the loop exit condition: that all queries + * have an accepted answer. */ + for (i=0; i<nqueries && alens[i]>0; i++); + if (i==nqueries) break; + if (t2-t1 >= retry_interval) { /* Query all configured namservers in parallel */ for (i=0; i<nqueries; i++) @@ -127,10 +194,20 @@ int __res_msend_rc(int nqueries, const unsigned char *const *queries, } /* Wait for a response, or until time to retry */ - if (poll(&pfd, 1, t1+retry_interval-t2) <= 0) continue; + if (poll(pfd, nqueries+1, t1+retry_interval-t2) <= 0) continue; - while ((rlen = recvfrom(fd, answers[next], asize, 0, - (void *)&sa, (socklen_t[1]){sl})) >= 0) { + while (next < nqueries) { + struct msghdr mh = { + .msg_name = (void *)&sa, + .msg_namelen = sl, + .msg_iovlen = 1, + .msg_iov = (struct iovec []){ + { .iov_base = (void *)answers[next], + .iov_len = asize } + } + }; + rlen = recvmsg(fd, &mh, 0); + if (rlen < 0) break; /* Ignore non-identifiable packets */ if (rlen < 4) continue; @@ -170,12 +247,72 @@ int __res_msend_rc(int nqueries, const unsigned char *const *queries, else memcpy(answers[i], answers[next], rlen); - if (next == nqueries) goto out; + /* Ignore further UDP if all slots full or TCP-mode */ + if (next == nqueries) pfd[nqueries].events = 0; + + /* If answer is truncated (TC bit), fallback to TCP */ + if ((answers[i][2] & 2) || (mh.msg_flags & MSG_TRUNC)) { + alens[i] = -1; + pthread_setcancelstate(PTHREAD_CANCEL_DISABLE, 0); + r = start_tcp(pfd+i, family, ns+j, sl, queries[i], qlens[i]); + pthread_setcancelstate(cs, 0); + if (r >= 0) { + qpos[i] = r; + apos[i] = 0; + } + continue; + } + } + + for (i=0; i<nqueries; i++) if (pfd[i].revents & POLLOUT) { + struct msghdr mh = { + .msg_iovlen = 2, + .msg_iov = (struct iovec [2]){ + { .iov_base = (uint8_t[]){ qlens[i]>>8, qlens[i] }, .iov_len = 2 }, + { .iov_base = (void *)queries[i], .iov_len = qlens[i] } } + }; + step_mh(&mh, qpos[i]); + r = sendmsg(pfd[i].fd, &mh, MSG_NOSIGNAL); + if (r < 0) goto out; + qpos[i] += r; + if (qpos[i] == qlens[i]+2) + pfd[i].events = POLLIN; + } + + for (i=0; i<nqueries; i++) if (pfd[i].revents & POLLIN) { + struct msghdr mh = { + .msg_iovlen = 2, + .msg_iov = (struct iovec [2]){ + { .iov_base = alen_buf[i], .iov_len = 2 }, + { .iov_base = answers[i], .iov_len = asize } } + }; + step_mh(&mh, apos[i]); + r = recvmsg(pfd[i].fd, &mh, 0); + if (r <= 0) goto out; + apos[i] += r; + if (apos[i] < 2) continue; + int alen = alen_buf[i][0]*256 + alen_buf[i][1]; + if (alen < 13) goto out; + if (apos[i] < alen+2 && apos[i] < asize+2) + continue; + int rcode = answers[i][3] & 15; + if (rcode != 0 && rcode != 3) + goto out; + + /* Storing the length here commits the accepted answer. + * Immediately close TCP socket so as not to consume + * resources we no longer need. */ + alens[i] = alen; + __syscall(SYS_close, pfd[i].fd); + pfd[i].fd = -1; } } out: pthread_cleanup_pop(1); + /* Disregard any incomplete TCP results */ + for (i=0; i<nqueries; i++) if (alens[i]<0) alens[i] = 0; + return 0; } diff --git a/src/network/res_query.c b/src/network/res_query.c index 2f4da2e2..506dc231 100644 --- a/src/network/res_query.c +++ b/src/network/res_query.c @@ -1,3 +1,4 @@ +#define _BSD_SOURCE #include <resolv.h> #include <netdb.h> @@ -6,7 +7,20 @@ int res_query(const char *name, int class, int type, unsigned char *dest, int le 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); + int r = __res_send(q, ql, dest, len); + if (r<12) { + h_errno = TRY_AGAIN; + return -1; + } + if ((dest[3] & 15) == 3) { + h_errno = HOST_NOT_FOUND; + return -1; + } + if ((dest[3] & 15) == 0 && !dest[6] && !dest[7]) { + h_errno = NO_DATA; + return -1; + } + return r; } weak_alias(res_query, res_search); diff --git a/src/network/res_send.c b/src/network/res_send.c index b9cea0bf..9593164d 100644 --- a/src/network/res_send.c +++ b/src/network/res_send.c @@ -1,9 +1,17 @@ #include <resolv.h> +#include <string.h> 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; + int r; + if (anslen < 512) { + unsigned char buf[512]; + r = __res_send(msg, msglen, buf, sizeof buf); + if (r >= 0) memcpy(answer, buf, r < anslen ? r : anslen); + return r; + } + r = __res_msend(1, &msg, &msglen, &answer, &anslen, anslen); + return r<0 || !anslen ? -1 : anslen; } weak_alias(__res_send, res_send); diff --git a/src/network/sendmsg.c b/src/network/sendmsg.c index 80cc5f41..acdfdf29 100644 --- a/src/network/sendmsg.c +++ b/src/network/sendmsg.c @@ -8,13 +8,16 @@ ssize_t sendmsg(int fd, const struct msghdr *msg, int flags) { #if LONG_MAX > INT_MAX struct msghdr h; - struct cmsghdr chbuf[1024/sizeof(struct cmsghdr)+1], *c; + /* Kernels before 2.6.38 set SCM_MAX_FD to 255, allocate enough + * space to support an SCM_RIGHTS ancillary message with 255 fds. + * Kernels since 2.6.38 set SCM_MAX_FD to 253. */ + struct cmsghdr chbuf[CMSG_SPACE(255*sizeof(int))/sizeof(struct cmsghdr)+1], *c; if (msg) { h = *msg; h.__pad1 = h.__pad2 = 0; msg = &h; if (h.msg_controllen) { - if (h.msg_controllen > 1024) { + if (h.msg_controllen > sizeof chbuf) { errno = ENOMEM; return -1; } diff --git a/src/network/setsockopt.c b/src/network/setsockopt.c index 2c188a96..612a1947 100644 --- a/src/network/setsockopt.c +++ b/src/network/setsockopt.c @@ -31,6 +31,15 @@ int setsockopt(int fd, int level, int optname, const void *optval, socklen_t opt r = __socketcall(setsockopt, fd, level, optname, ((long[]){s, CLAMP(us)}), 2*sizeof(long), 0); + break; + case SO_TIMESTAMP: + case SO_TIMESTAMPNS: + if (SO_TIMESTAMP == SO_TIMESTAMP_OLD) break; + if (optname==SO_TIMESTAMP) optname=SO_TIMESTAMP_OLD; + if (optname==SO_TIMESTAMPNS) optname=SO_TIMESTAMPNS_OLD; + r = __socketcall(setsockopt, fd, level, + optname, optval, optlen, 0); + break; } } return __syscall_ret(r); diff --git a/src/network/socket.c b/src/network/socket.c index a2e92d90..afa1a7f3 100644 --- a/src/network/socket.c +++ b/src/network/socket.c @@ -5,17 +5,17 @@ int socket(int domain, int type, int protocol) { - int s = socketcall(socket, domain, type, protocol, 0, 0, 0); - if (s<0 && (errno==EINVAL || errno==EPROTONOSUPPORT) + int s = __socketcall(socket, domain, type, protocol, 0, 0, 0); + if ((s==-EINVAL || s==-EPROTONOSUPPORT) && (type&(SOCK_CLOEXEC|SOCK_NONBLOCK))) { - s = socketcall(socket, domain, + s = __socketcall(socket, domain, type & ~(SOCK_CLOEXEC|SOCK_NONBLOCK), protocol, 0, 0, 0); - if (s < 0) return s; + if (s < 0) return __syscall_ret(s); if (type & SOCK_CLOEXEC) __syscall(SYS_fcntl, s, F_SETFD, FD_CLOEXEC); if (type & SOCK_NONBLOCK) __syscall(SYS_fcntl, s, F_SETFL, O_NONBLOCK); } - return s; + return __syscall_ret(s); } |