#include "secnet.h"
#include "util.h"
#include "netlink.h"
+#include "process.h"
+#include "unaligned.h"
#include <stdio.h>
#include <string.h>
#include <unistd.h>
+#include <errno.h>
+#include <fcntl.h>
#define SLIP_END 192
#define SLIP_ESC 219
#define SLIP_ESCEND 220
#define SLIP_ESCESC 221
-/* Connection to the kernel through userv-ipif */
-
-struct userv {
+struct slip {
struct netlink nl;
- int txfd; /* We transmit to userv */
- int rxfd; /* We receive from userv */
- string_t userv_path;
- string_t service_user;
- string_t service_name;
- uint32_t txbuflen;
struct buffer_if *buff; /* We unstuff received packets into here
- and send them to the site code. */
+ and send them to the netlink code. */
bool_t pending_esc;
+ bool_t ignoring_packet; /* If this packet was corrupt or overlong,
+ we ignore everything up to the next END */
netlink_deliver_fn *netlink_to_tunnel;
- uint32_t local_address; /* host interface address */
};
-static int userv_beforepoll(void *sst, struct pollfd *fds, int *nfds_io,
- int *timeout_io, const struct timeval *tv_now,
- uint64_t *now)
-{
- struct userv *st=sst;
- *nfds_io=2;
- fds[0].fd=st->txfd;
- fds[0].events=POLLERR; /* Might want to pick up POLLOUT sometime */
- fds[1].fd=st->rxfd;
- fds[1].events=POLLIN|POLLERR|POLLHUP;
- return 0;
-}
-
-static void userv_afterpoll(void *sst, struct pollfd *fds, int nfds,
- const struct timeval *tv_now, uint64_t *now)
-{
- struct userv *st=sst;
- uint8_t rxbuf[DEFAULT_BUFSIZE];
- int l,i;
-
- if (fds[1].revents&POLLERR) {
- Message(M_ERROR,"%s: userv_afterpoll: hup!\n",st->nl.name);
- }
- if (fds[1].revents&POLLIN) {
- l=read(st->rxfd,rxbuf,DEFAULT_BUFSIZE);
- if (l<0) {
- fatal_perror("%s: userv_afterpoll: read(rxfd)",st->nl.name);
- }
- if (l==0) {
- fatal("%s: userv_afterpoll: read(rxfd)=0; userv gone away?\n",
- st->nl.name);
- }
- /* XXX really crude unstuff code */
- /* XXX check for buffer overflow */
- BUF_ASSERT_USED(st->buff);
- for (i=0; i<l; i++) {
- if (st->pending_esc) {
- st->pending_esc=False;
- switch(rxbuf[i]) {
- case SLIP_ESCEND:
- *(uint8_t *)buf_append(st->buff,1)=SLIP_END;
- break;
- case SLIP_ESCESC:
- *(uint8_t *)buf_append(st->buff,1)=SLIP_ESC;
- break;
- default:
- fatal("userv_afterpoll: bad SLIP escape character\n");
- }
- } else {
- switch (rxbuf[i]) {
- case SLIP_END:
- if (st->buff->size>0) {
- st->netlink_to_tunnel(&st->nl,NULL,
- st->buff);
- BUF_ALLOC(st->buff,"userv_afterpoll");
- }
- buffer_init(st->buff,st->nl.max_start_pad);
- break;
- case SLIP_ESC:
- st->pending_esc=True;
- break;
- default:
- *(uint8_t *)buf_append(st->buff,1)=rxbuf[i];
- break;
- }
- }
- }
- }
-}
+/* Generic SLIP mangling code */
-/* Send buf to the kernel. Free buf before returning. */
-static void userv_deliver_to_kernel(void *sst, void *cid,
- struct buffer_if *buf)
+static void slip_stuff(struct slip *st, struct buffer_if *buf, int fd)
{
- struct userv *st=sst;
uint8_t txbuf[DEFAULT_BUFSIZE];
uint8_t *i;
- uint32_t j;
+ int32_t j=0;
BUF_ASSERT_USED(buf);
- /* Spit the packet at userv-ipif: SLIP start marker, then
- bytestuff the packet, then SLIP end marker */
- /* XXX crunchy bytestuff code */
- j=0;
+ /* There's probably a much more efficient way of implementing this */
txbuf[j++]=SLIP_END;
for (i=buf->start; i<(buf->start+buf->size); i++) {
switch (*i) {
txbuf[j++]=*i;
break;
}
+ if ((j+2)>DEFAULT_BUFSIZE) {
+ if (write(fd,txbuf,j)<0) {
+ fatal_perror("slip_stuff: write()");
+ }
+ j=0;
+ }
}
txbuf[j++]=SLIP_END;
- if (write(st->txfd,txbuf,j)<0) {
- fatal_perror("userv_deliver_to_kernel: write()");
+ if (write(fd,txbuf,j)<0) {
+ fatal_perror("slip_stuff: write()");
}
BUF_FREE(buf);
}
-static void userv_phase_hook(void *sst, uint32_t newphase)
+static void slip_unstuff(struct slip *st, uint8_t *buf, uint32_t l)
+{
+ uint32_t i;
+
+ BUF_ASSERT_USED(st->buff);
+ for (i=0; i<l; i++) {
+ int outputchr;
+ enum { OUTPUT_END = 256, OUTPUT_NOTHING = 257 };
+
+ if (!st->buff->size)
+ buffer_init(st->buff,calculate_max_start_pad());
+
+ if (st->pending_esc) {
+ st->pending_esc=False;
+ switch(buf[i]) {
+ case SLIP_ESCEND:
+ outputchr=SLIP_END;
+ break;
+ case SLIP_ESCESC:
+ outputchr=SLIP_ESC;
+ break;
+ default:
+ if (!st->ignoring_packet) {
+ Message(M_WARNING, "userv_afterpoll: bad SLIP escape"
+ " character, dropping packet\n");
+ }
+ st->ignoring_packet=True;
+ outputchr=OUTPUT_NOTHING;
+ break;
+ }
+ } else {
+ switch (buf[i]) {
+ case SLIP_END:
+ outputchr=OUTPUT_END;
+ break;
+ case SLIP_ESC:
+ st->pending_esc=True;
+ outputchr=OUTPUT_NOTHING;
+ break;
+ default:
+ outputchr=buf[i];
+ break;
+ }
+ }
+
+ if (st->ignoring_packet) {
+ if (outputchr == OUTPUT_END) {
+ st->ignoring_packet=False;
+ st->buff->size=0;
+ }
+ } else {
+ if (outputchr == OUTPUT_END) {
+ if (st->buff->size>0) {
+ st->netlink_to_tunnel(&st->nl,st->buff);
+ BUF_ALLOC(st->buff,"userv_afterpoll");
+ }
+ st->buff->size=0;
+ } else if (outputchr != OUTPUT_NOTHING) {
+ if (buf_remaining_space(st->buff)) {
+ buf_append_uint8(st->buff,outputchr);
+ } else {
+ Message(M_WARNING, "userv_afterpoll: dropping overlong"
+ " SLIP packet\n");
+ st->ignoring_packet=True;
+ }
+ }
+ }
+ }
+}
+
+static void slip_init(struct slip *st, struct cloc loc, dict_t *dict,
+ cstring_t name, netlink_deliver_fn *to_host)
+{
+ st->netlink_to_tunnel=
+ netlink_init(&st->nl,st,loc,dict,
+ "netlink-userv-ipif",NULL,to_host);
+ st->buff=find_cl_if(dict,"buffer",CL_BUFFER,True,"name",loc);
+ BUF_ALLOC(st->buff,"slip_init");
+ st->pending_esc=False;
+ st->ignoring_packet=False;
+}
+
+/* Connection to the kernel through userv-ipif */
+
+struct userv {
+ struct slip slip;
+ int txfd; /* We transmit to userv */
+ int rxfd; /* We receive from userv */
+ cstring_t userv_path;
+ cstring_t service_user;
+ cstring_t service_name;
+ pid_t pid;
+ bool_t expecting_userv_exit;
+};
+
+static int userv_beforepoll(void *sst, struct pollfd *fds, int *nfds_io,
+ int *timeout_io)
+{
+ struct userv *st=sst;
+
+ if (st->rxfd!=-1) {
+ BEFOREPOLL_WANT_FDS(2);
+ fds[0].fd=st->txfd;
+ fds[0].events=0; /* Might want to pick up POLLOUT sometime */
+ fds[1].fd=st->rxfd;
+ fds[1].events=POLLIN;
+ } else {
+ BEFOREPOLL_WANT_FDS(0);
+ }
+ return 0;
+}
+
+static void userv_afterpoll(void *sst, struct pollfd *fds, int nfds)
+{
+ struct userv *st=sst;
+ uint8_t rxbuf[DEFAULT_BUFSIZE];
+ int l;
+
+ if (nfds==0) return;
+
+ if (fds[1].revents&POLLERR) {
+ Message(M_ERR,"%s: userv_afterpoll: POLLERR!\n",st->slip.nl.name);
+ }
+ if (fds[1].revents&POLLIN) {
+ l=read(st->rxfd,rxbuf,DEFAULT_BUFSIZE);
+ if (l<0) {
+ if (errno!=EINTR)
+ fatal_perror("%s: userv_afterpoll: read(rxfd)",
+ st->slip.nl.name);
+ } else if (l==0) {
+ fatal("%s: userv_afterpoll: read(rxfd)=0; userv gone away?",
+ st->slip.nl.name);
+ } else slip_unstuff(&st->slip,rxbuf,l);
+ }
+}
+
+/* Send buf to the kernel. Free buf before returning. */
+static void userv_deliver_to_kernel(void *sst, struct buffer_if *buf)
+{
+ struct userv *st=sst;
+
+ if (buf->size > st->slip.nl.mtu) {
+ Message(M_ERR,"%s: packet of size %"PRIu32" exceeds mtu %"PRIu32":"
+ " cannot be injected into kernel, dropped\n",
+ st->slip.nl.name, buf->size, st->slip.nl.mtu);
+ BUF_FREE(buf);
+ return;
+ }
+
+ slip_stuff(&st->slip,buf,st->txfd);
+}
+
+static void userv_userv_callback(void *sst, pid_t pid, int status)
{
struct userv *st=sst;
- pid_t child;
+
+ if (pid!=st->pid) {
+ Message(M_WARNING,"userv_callback called unexpectedly with pid %d "
+ "(expected %d)\n",pid,st->pid);
+ return;
+ }
+ if (!st->expecting_userv_exit) {
+ if (WIFEXITED(status)) {
+ fatal("%s: userv exited unexpectedly with status %d",
+ st->slip.nl.name,WEXITSTATUS(status));
+ } else if (WIFSIGNALED(status)) {
+ fatal("%s: userv exited unexpectedly: uncaught signal %d",
+ st->slip.nl.name,WTERMSIG(status));
+ } else {
+ fatal("%s: userv stopped unexpectedly",
+ st->slip.nl.name);
+ }
+ }
+ Message(M_WARNING,"%s: userv subprocess died with status %d\n",
+ st->slip.nl.name,WEXITSTATUS(status));
+ st->pid=0;
+}
+
+struct userv_entry_rec {
+ cstring_t path;
+ const char **argv;
+ int in;
+ int out;
+ /* XXX perhaps we should collect and log stderr? */
+};
+
+static void userv_entry(void *sst)
+{
+ struct userv_entry_rec *st=sst;
+
+ dup2(st->in,0);
+ dup2(st->out,1);
+
+ setsid();
+ /* XXX We really should strdup() all of argv[] but because we'll just
+ exit anyway if execvp() fails it doesn't seem worth bothering. */
+ execvp(st->path,(char *const*)st->argv);
+ perror("userv-entry: execvp()");
+ exit(1);
+}
+
+static void userv_invoke_userv(struct userv *st)
+{
+ struct userv_entry_rec er[1];
int c_stdin[2];
int c_stdout[2];
- string_t addrs;
string_t nets;
string_t s;
- struct netlink_route *r;
- int i;
+ struct netlink_client *r;
+ struct ipset *allnets;
+ struct subnet_list *snets;
+ int i, nread;
+ uint8_t confirm;
+
+ if (st->pid) {
+ fatal("userv_invoke_userv: already running");
+ }
/* This is where we actually invoke userv - all the networks we'll
be using should already have been registered. */
- addrs=safe_malloc(512,"userv_phase_hook:addrs");
- snprintf(addrs,512,"%s,%s,%d,slip",ipaddr_to_string(st->local_address),
- ipaddr_to_string(st->nl.secnet_address),st->nl.mtu);
+ char addrs[512];
+ snprintf(addrs,sizeof(addrs),"%s,%s,%d,slip",
+ ipaddr_to_string(st->slip.nl.local_address),
+ ipaddr_to_string(st->slip.nl.secnet_address),st->slip.nl.mtu);
- nets=safe_malloc(1024,"userv_phase_hook:nets");
- *nets=0;
- r=st->nl.routes;
- for (i=0; i<st->nl.n_routes; i++) {
- if (r[i].up) {
- r[i].kup=True;
- s=subnet_to_string(&r[i].net);
- strcat(nets,s);
- strcat(nets,",");
- free(s);
+ allnets=ipset_new();
+ for (r=st->slip.nl.clients; r; r=r->next) {
+ if (r->link_quality > LINK_QUALITY_UNUSED) {
+ struct ipset *nan;
+ r->kup=True;
+ nan=ipset_union(allnets,r->networks);
+ ipset_free(allnets);
+ allnets=nan;
}
}
+ snets=ipset_to_subnet_list(allnets);
+ ipset_free(allnets);
+ nets=safe_malloc(20*snets->entries,"userv_invoke_userv:nets");
+ *nets=0;
+ for (i=0; i<snets->entries; i++) {
+ s=subnet_to_string(snets->list[i]);
+ strcat(nets,s);
+ strcat(nets,",");
+ }
nets[strlen(nets)-1]=0;
+ subnet_list_free(snets);
- Message(M_INFO,"%s: about to invoke: %s %s %s %s %s\n",st->nl.name,
+ Message(M_INFO,"%s: about to invoke: %s %s %s %s %s\n",st->slip.nl.name,
st->userv_path,st->service_user,st->service_name,addrs,nets);
- /* Allocate buffer, plus space for padding. Make sure we end up
- with the start of the packet well-aligned. */
- /* ALIGN(st->max_start_pad,16); */
- /* ALIGN(st->max_end_pad,16); */
-
- st->pending_esc=False;
+ st->slip.pending_esc=False;
/* Invoke userv */
- if (pipe(c_stdin)!=0) {
- fatal_perror("userv_phase_hook: pipe(c_stdin)");
- }
- if (pipe(c_stdout)!=0) {
- fatal_perror("userv_phase_hook: pipe(c_stdout)");
- }
+ pipe_cloexec(c_stdin);
+ pipe_cloexec(c_stdout);
st->txfd=c_stdin[1];
st->rxfd=c_stdout[0];
- child=fork();
- if (child==-1) {
- fatal_perror("userv_phase_hook: fork()");
+ er->in=c_stdin[0];
+ er->out=c_stdout[1];
+ /* The arguments are:
+ userv
+ service-user
+ service-name
+ local-addr,secnet-addr,mtu,protocol
+ route1,route2,... */
+ const char *er_argv[6];
+ er->argv=er_argv;
+ er->argv[0]=st->userv_path;
+ er->argv[1]=st->service_user;
+ er->argv[2]=st->service_name;
+ er->argv[3]=addrs;
+ er->argv[4]=nets;
+ er->argv[5]=NULL;
+ er->path=st->userv_path;
+
+ st->pid=makesubproc(userv_entry, userv_userv_callback,
+ er, st, st->slip.nl.name);
+ close(er->in);
+ close(er->out);
+ free(nets);
+ Message(M_INFO,"%s: userv-ipif pid is %d\n",st->slip.nl.name,st->pid);
+ /* Read a single character from the pipe to confirm userv-ipif is
+ running. If we get a SIGCHLD at this point then we'll get EINTR. */
+ if ((nread=read(st->rxfd,&confirm,1))!=1) {
+ if (errno==EINTR) {
+ Message(M_WARNING,"%s: read of confirmation byte was "
+ "interrupted\n",st->slip.nl.name);
+ } else {
+ if (nread<0) {
+ fatal_perror("%s: error reading confirmation byte",
+ st->slip.nl.name);
+ } else {
+ fatal("%s: unexpected EOF instead of confirmation byte"
+ " - userv ipif failed?", st->slip.nl.name);
+ }
+ }
+ } else {
+ if (confirm!=SLIP_END) {
+ fatal("%s: bad confirmation byte %d from userv-ipif",
+ st->slip.nl.name,confirm);
+ }
+ }
+}
+
+static void userv_kill_userv(struct userv *st)
+{
+ if (st->pid) {
+ kill(-st->pid,SIGTERM);
+ st->expecting_userv_exit=True;
+ }
+}
+
+static void userv_phase_hook(void *sst, uint32_t newphase)
+{
+ struct userv *st=sst;
+ /* We must wait until signal processing has started before forking
+ userv */
+ if (newphase==PHASE_RUN) {
+ userv_invoke_userv(st);
+ /* Register for poll() */
+ register_for_poll(st, userv_beforepoll, userv_afterpoll,
+ st->slip.nl.name);
}
- if (child==0) {
- char **argv;
-
- /* We are the child. Modify our stdin and stdout, then exec userv */
- dup2(c_stdin[0],0);
- dup2(c_stdout[1],1);
- close(c_stdin[1]);
- close(c_stdout[0]);
-
- /* The arguments are:
- userv
- service-user
- service-name
- local-addr,secnet-addr,mtu,protocol
- route1,route2,... */
- argv=malloc(sizeof(*argv)*6);
- argv[0]=st->userv_path;
- argv[1]=st->service_user;
- argv[2]=st->service_name;
- argv[3]=addrs;
- argv[4]=nets;
- argv[5]=NULL;
- execvp(st->userv_path,argv);
- perror("netlink-userv-ipif: execvp");
-
- exit(1);
+ if (newphase==PHASE_SHUTDOWN) {
+ userv_kill_userv(st);
}
- /* We are the parent... */
-
- /* Register for poll() */
- register_for_poll(st, userv_beforepoll, userv_afterpoll, 2, st->nl.name);
}
static list_t *userv_apply(closure_t *self, struct cloc loc, dict_t *context,
dict=item->data.dict;
- st->netlink_to_tunnel=
- netlink_init(&st->nl,st,loc,dict,
- "netlink-userv-ipif",NULL,userv_deliver_to_kernel);
+ slip_init(&st->slip,loc,dict,"netlink-userv-ipif",
+ userv_deliver_to_kernel);
st->userv_path=dict_read_string(dict,"userv-path",False,"userv-netlink",
loc);
if (!st->userv_path) st->userv_path="userv";
if (!st->service_user) st->service_user="root";
if (!st->service_name) st->service_name="ipif";
- st->buff=find_cl_if(dict,"buffer",CL_BUFFER,True,"userv-netlink",loc);
- st->local_address=string_to_ipaddr(
- dict_find_item(dict,"local-address", True, "netlink", loc),"netlink");
- BUF_ALLOC(st->buff,"netlink:userv_apply");
-
st->rxfd=-1; st->txfd=-1;
- add_hook(PHASE_DROPPRIV,userv_phase_hook,st);
+ st->pid=0;
+ st->expecting_userv_exit=False;
+ add_hook(PHASE_RUN,userv_phase_hook,st);
+ add_hook(PHASE_SHUTDOWN,userv_phase_hook,st);
- return new_closure(&st->nl.cl);
+ return new_closure(&st->slip.nl.cl);
}
-init_module slip_module;
void slip_module(dict_t *dict)
{
add_closure(dict,"userv-ipif",userv_apply);
-#if 0
- /* TODO */
- add_closure(dict,"pty-slip",ptyslip_apply);
- add_closure(dict,"slipd",slipd_apply);
-#endif /* 0 */
}