a200e4f865
*** THIS RFC INCLUDES A MINOR CHANGE TO THE MPI-RTE INTERFACE *** Note: during the course of this work, it was necessary to completely separate the MPI and RTE progress engines. There were multiple places in the MPI layer where ORTE_WAIT_FOR_COMPLETION was being used. A new OMPI_WAIT_FOR_COMPLETION macro was created (defined in ompi/mca/rte/rte.h) that simply cycles across opal_progress until the provided flag becomes false. Places where the MPI layer blocked waiting for RTE to complete an event have been modified to use this macro. *************************************************************************************** I am reissuing this RFC because of the time that has passed since its original release. Since its initial release and review, I have debugged it further to ensure it fully supports tests like loop_spawn. It therefore seems ready for merge back to the trunk. Given its prior review, I have set the timeout for one week. The code is in https://bitbucket.org/rhc/ompi-oob2 WHAT: Rewrite of ORTE OOB WHY: Support asynchronous progress and a host of other features WHEN: Wed, August 21 SYNOPSIS: The current OOB has served us well, but a number of limitations have been identified over the years. Specifically: * it is only progressed when called via opal_progress, which can lead to hangs or recursive calls into libevent (which is not supported by that code) * we've had issues when multiple NICs are available as the code doesn't "shift" messages between transports - thus, all nodes had to be available via the same TCP interface. * the OOB "unloads" incoming opal_buffer_t objects during the transmission, thus preventing use of OBJ_RETAIN in the code when repeatedly sending the same message to multiple recipients * there is no failover mechanism across NICs - if the selected NIC (or its attached switch) fails, we are forced to abort * only one transport (i.e., component) can be "active" The revised OOB resolves these problems: * async progress is used for all application processes, with the progress thread blocking in the event library * each available TCP NIC is supported by its own TCP module. The ability to asynchronously progress each module independently is provided, but not enabled by default (a runtime MCA parameter turns it "on") * multi-address TCP NICs (e.g., a NIC with both an IPv4 and IPv6 address, or with virtual interfaces) are supported - reachability is determined by comparing the contact info for a peer against all addresses within the range covered by the address/mask pairs for the NIC. * a message that arrives on one TCP NIC is automatically shifted to whatever NIC that is connected to the next "hop" if that peer cannot be reached by the incoming NIC. If no TCP module will reach the peer, then the OOB attempts to send the message via all other available components - if none can reach the peer, then an "error" is reported back to the RML, which then calls the errmgr for instructions. * opal_buffer_t now conforms to standard object rules re OBJ_RETAIN as we no longer "unload" the incoming object * NIC failure is reported to the TCP component, which then tries to resend the message across any other available TCP NIC. If that doesn't work, then the message is given back to the OOB base to try using other components. If all that fails, then the error is reported to the RML, which reports to the errmgr for instructions * obviously from the above, multiple OOB components (e.g., TCP and UD) can be active in parallel * the matching code has been moved to the RML (and out of the OOB/TCP component) so it is independent of transport * routing is done by the individual OOB modules (as opposed to the RML). Thus, both routed and non-routed transports can simultaneously be active * all blocking send/recv APIs have been removed. Everything operates asynchronously. KNOWN LIMITATIONS: * although provision is made for component failover as described above, the code for doing so has not been fully implemented yet. At the moment, if all connections for a given peer fail, the errmgr is notified of a "lost connection", which by default results in termination of the job if it was a lifeline * the IPv6 code is present and compiles, but is not complete. Since the current IPv6 support in the OOB doesn't work anyway, I don't consider this a blocker * routing is performed at the individual module level, yet the active routed component is selected on a global basis. We probably should update that to reflect that different transports may need/choose to route in different ways * obviously, not every error path has been tested nor necessarily covered * determining abnormal termination is more challenging than in the old code as we now potentially have multiple ways of connecting to a process. Ideally, we would declare "connection failed" when *all* transports can no longer reach the process, but that requires some additional (possibly complex) code. For now, the code replicates the old behavior only somewhat modified - i.e., if a module sees its connection fail, it checks to see if it is a lifeline. If so, it notifies the errmgr that the lifeline is lost - otherwise, it notifies the errmgr that a non-lifeline connection was lost. * reachability is determined solely on the basis of a shared subnet address/mask - more sophisticated algorithms (e.g., the one used in the tcp btl) are required to handle routing via gateways * the RML needs to assign sequence numbers to each message on a per-peer basis. The receiving RML will then deliver messages in order, thus preventing out-of-order messaging in the case where messages travel across different transports or a message needs to be redirected/resent due to failure of a NIC This commit was SVN r29058.
232 строки
7.5 KiB
C
232 строки
7.5 KiB
C
/*
|
|
* Copyright (c) 2010 Cisco Systems, Inc. All rights reserved.
|
|
* Copyright (c) 2010 Oracle and/or its affiliates. All rights reserved.
|
|
* $COPYRIGHT$
|
|
*
|
|
* Additional copyrights may follow
|
|
*
|
|
* $HEADER$
|
|
*/
|
|
|
|
#include "opal_config.h"
|
|
#include "opal/constants.h"
|
|
#include "opal/util/output.h"
|
|
|
|
#include <string.h>
|
|
#ifdef HAVE_UNISTD_H
|
|
#include <unistd.h>
|
|
#endif
|
|
#include <errno.h>
|
|
#ifdef HAVE_SYS_TYPES_H
|
|
#include <sys/types.h>
|
|
#endif
|
|
#ifdef HAVE_SYS_SOCKET_H
|
|
#include <sys/socket.h>
|
|
#endif
|
|
#ifdef HAVE_SYS_SOCKIO_H
|
|
#include <sys/sockio.h>
|
|
#endif
|
|
#ifdef HAVE_SYS_IOCTL_H
|
|
#include <sys/ioctl.h>
|
|
#endif
|
|
#ifdef HAVE_NETINET_IN_H
|
|
#include <netinet/in.h>
|
|
#endif
|
|
#ifdef HAVE_ARPA_INET_H
|
|
#include <arpa/inet.h>
|
|
#endif
|
|
#ifdef HAVE_NET_IF_H
|
|
#if defined(__APPLE__) && defined(_LP64)
|
|
/* Apple engineering suggested using options align=power as a
|
|
workaround for a bug in OS X 10.4 (Tiger) that prevented ioctl(...,
|
|
SIOCGIFCONF, ...) from working properly in 64 bit mode on Power PC.
|
|
It turns out that the underlying issue is the size of struct
|
|
ifconf, which the kernel expects to be 12 and natural 64 bit
|
|
alignment would make 16. The same bug appears in 64 bit mode on
|
|
Intel macs, but align=power is a no-op there, so instead, use the
|
|
pack pragma to instruct the compiler to pack on 4 byte words, which
|
|
has the same effect as align=power for our needs and works on both
|
|
Intel and Power PC Macs. */
|
|
#pragma pack(push,4)
|
|
#endif
|
|
#include <net/if.h>
|
|
#if defined(__APPLE__) && defined(_LP64)
|
|
#pragma pack(pop)
|
|
#endif
|
|
#endif
|
|
#ifdef HAVE_NETDB_H
|
|
#include <netdb.h>
|
|
#endif
|
|
#ifdef HAVE_IFADDRS_H
|
|
#include <ifaddrs.h>
|
|
#endif
|
|
|
|
#include "opal/mca/if/if.h"
|
|
#include "opal/mca/if/base/base.h"
|
|
|
|
static int if_bsdx_ipv6_open(void);
|
|
|
|
/* Discovers IPv6 interfaces for:
|
|
*
|
|
* NetBSD
|
|
* OpenBSD
|
|
* FreeBSD
|
|
* 386BSD
|
|
* bsdi
|
|
* Apple
|
|
*/
|
|
opal_if_base_component_t mca_if_bsdx_ipv6_component = {
|
|
/* First, the mca_component_t struct containing meta information
|
|
about the component itself */
|
|
{
|
|
OPAL_IF_BASE_VERSION_2_0_0,
|
|
|
|
/* Component name and version */
|
|
"bsdx_ipv6",
|
|
OPAL_MAJOR_VERSION,
|
|
OPAL_MINOR_VERSION,
|
|
OPAL_RELEASE_VERSION,
|
|
|
|
/* Component open and close functions */
|
|
if_bsdx_ipv6_open,
|
|
NULL
|
|
},
|
|
{
|
|
/* This component is checkpointable */
|
|
MCA_BASE_METADATA_PARAM_CHECKPOINT
|
|
},
|
|
};
|
|
|
|
/* configure using getifaddrs(3) */
|
|
static int if_bsdx_ipv6_open(void)
|
|
{
|
|
#if OPAL_ENABLE_IPV6
|
|
struct ifaddrs **ifadd_list;
|
|
struct ifaddrs *cur_ifaddrs;
|
|
struct sockaddr_in6* sin_addr;
|
|
|
|
opal_output_verbose(1, opal_if_base_framework.framework_output,
|
|
"searching for IPv6 interfaces");
|
|
|
|
/*
|
|
* the manpage claims that getifaddrs() allocates the memory,
|
|
* and freeifaddrs() is later used to release the allocated memory.
|
|
* however, without this malloc the call to getifaddrs() segfaults
|
|
*/
|
|
ifadd_list = (struct ifaddrs **) malloc(sizeof(struct ifaddrs*));
|
|
|
|
/* create the linked list of ifaddrs structs */
|
|
if (getifaddrs(ifadd_list) < 0) {
|
|
opal_output(0, "opal_ifinit: getifaddrs() failed with error=%d\n",
|
|
errno);
|
|
free(ifadd_list);
|
|
return OPAL_ERROR;
|
|
}
|
|
|
|
for (cur_ifaddrs = *ifadd_list; NULL != cur_ifaddrs;
|
|
cur_ifaddrs = cur_ifaddrs->ifa_next) {
|
|
opal_if_t *intf;
|
|
struct in6_addr a6;
|
|
|
|
/* skip non-ipv6 interface addresses */
|
|
if (AF_INET6 != cur_ifaddrs->ifa_addr->sa_family) {
|
|
opal_output_verbose(1, opal_if_base_framework.framework_output,
|
|
"skipping non-ipv6 interface %s[%d].\n",
|
|
cur_ifaddrs->ifa_name, (int)cur_ifaddrs->ifa_addr->sa_family);
|
|
continue;
|
|
}
|
|
|
|
/* skip interface if it is down (IFF_UP not set) */
|
|
if (0 == (cur_ifaddrs->ifa_flags & IFF_UP)) {
|
|
opal_output_verbose(1, opal_if_base_framework.framework_output,
|
|
"skipping non-up interface %s.\n", cur_ifaddrs->ifa_name);
|
|
continue;
|
|
}
|
|
|
|
/* skip interface if it is a loopback device (IFF_LOOPBACK set) */
|
|
if (!opal_if_retain_loopback && 0 != (cur_ifaddrs->ifa_flags & IFF_LOOPBACK)) {
|
|
opal_output_verbose(1, opal_if_base_framework.framework_output,
|
|
"skipping loopback interface %s.\n", cur_ifaddrs->ifa_name);
|
|
continue;
|
|
}
|
|
|
|
/* or if it is a point-to-point interface */
|
|
/* TODO: do we really skip p2p? */
|
|
if (0!= (cur_ifaddrs->ifa_flags & IFF_POINTOPOINT)) {
|
|
opal_output_verbose(1, opal_if_base_framework.framework_output,
|
|
"skipping p2p interface %s.\n", cur_ifaddrs->ifa_name);
|
|
continue;
|
|
}
|
|
|
|
sin_addr = (struct sockaddr_in6 *) cur_ifaddrs->ifa_addr;
|
|
|
|
/*
|
|
* skip IPv6 address starting with fe80:, as this is supposed to be
|
|
* link-local scope. sockaddr_in6->sin6_scope_id doesn't always work
|
|
* TODO: test whether scope id is set to a sensible value on
|
|
* linux and/or bsd (including osx)
|
|
*
|
|
* MacOSX: fe80::... has a scope of 0, but ifconfig -a shows
|
|
* a scope of 4 on that particular machine,
|
|
* so the scope returned by getifaddrs() isn't working properly
|
|
*/
|
|
|
|
if ((IN6_IS_ADDR_LINKLOCAL (&sin_addr->sin6_addr))) {
|
|
opal_output_verbose(1, opal_if_base_framework.framework_output,
|
|
"skipping link-local ipv6 address on interface "
|
|
"%s with scope %d.\n",
|
|
cur_ifaddrs->ifa_name, sin_addr->sin6_scope_id);
|
|
continue;
|
|
}
|
|
|
|
if (0 < opal_output_get_verbosity(opal_if_base_framework.framework_output)) {
|
|
char *addr_name = (char *) malloc(48*sizeof(char));
|
|
inet_ntop(AF_INET6, &sin_addr->sin6_addr, addr_name, 48*sizeof(char));
|
|
opal_output(0, "ipv6 capable interface %s discovered, address %s.\n",
|
|
cur_ifaddrs->ifa_name, addr_name);
|
|
free(addr_name);
|
|
}
|
|
|
|
/* fill values into the opal_if_t */
|
|
memcpy(&a6, &(sin_addr->sin6_addr), sizeof(struct in6_addr));
|
|
|
|
intf = OBJ_NEW(opal_if_t);
|
|
if (NULL == intf) {
|
|
opal_output(0, "opal_ifinit: unable to allocate %lu bytes\n",
|
|
sizeof(opal_if_t));
|
|
free(ifadd_list);
|
|
return OPAL_ERR_OUT_OF_RESOURCE;
|
|
}
|
|
intf->af_family = AF_INET6;
|
|
strncpy(intf->if_name, cur_ifaddrs->ifa_name, IF_NAMESIZE);
|
|
intf->if_index = opal_list_get_size(&opal_if_list) + 1;
|
|
((struct sockaddr_in6*) &intf->if_addr)->sin6_addr = a6;
|
|
((struct sockaddr_in6*) &intf->if_addr)->sin6_family = AF_INET6;
|
|
|
|
/* since every scope != 0 is ignored, we just set the scope to 0 */
|
|
((struct sockaddr_in6*) &intf->if_addr)->sin6_scope_id = 0;
|
|
|
|
/*
|
|
* hardcoded netmask, adrian says that's ok
|
|
*/
|
|
intf->if_mask = 64;
|
|
intf->if_flags = cur_ifaddrs->ifa_flags;
|
|
|
|
/*
|
|
* FIXME: figure out how to gain access to the kernel index
|
|
* (or create our own), getifaddrs() does not contain such
|
|
* data
|
|
*/
|
|
intf->if_kernel_index =
|
|
(uint16_t) if_nametoindex(cur_ifaddrs->ifa_name);
|
|
opal_list_append(&opal_if_list, &(intf->super));
|
|
} /* of for loop over ifaddrs list */
|
|
|
|
free(ifadd_list);
|
|
#endif /* OPAL_ENABLE_IPV6 */
|
|
|
|
return OPAL_SUCCESS;
|
|
}
|
|
|
|
|