| Message ID | 20180301125846.5385-1-yuval.shaia@oracle.com (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Leon Romanovsky |
| Headers | show |
On Thu, Mar 01, 2018 at 02:58:46PM +0200, Yuval Shaia wrote: > Add a command line option to allow selecting specific address as done in > rdma_client. > > Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com> > --- > librdmacm/examples/rdma_server.c | 8 ++++++-- > 1 file changed, 6 insertions(+), 2 deletions(-) > > diff --git a/librdmacm/examples/rdma_server.c b/librdmacm/examples/rdma_server.c > index bcf649fe..77c021ac 100644 > --- a/librdmacm/examples/rdma_server.c > +++ b/librdmacm/examples/rdma_server.c > @@ -36,6 +36,7 @@ > #include <rdma/rdma_cma.h> > #include <rdma/rdma_verbs.h> > > +static const char *server = "127.0.0.1"; > static const char *port = "7471"; > > static struct rdma_cm_id *listen_id, *id; > @@ -55,7 +56,7 @@ static int run(void) > memset(&hints, 0, sizeof hints); > hints.ai_flags = RAI_PASSIVE; > hints.ai_port_space = RDMA_PS_TCP; > - ret = rdma_getaddrinfo(NULL, port, &hints, &res); > + ret = rdma_getaddrinfo(server, port, &hints, &res); In such case (server is going to be not-NULL), the AI_PASSIVE hint will be ignored by getaddrinfo, but don't know whenever it is safe to remove "hints.ai_flags = RAI_PASSIVE;" line above. What do you think? Thanks
On Mon, Mar 05, 2018 at 08:55:25AM +0200, Leon Romanovsky wrote: > On Thu, Mar 01, 2018 at 02:58:46PM +0200, Yuval Shaia wrote: > > Add a command line option to allow selecting specific address as done in > > rdma_client. > > > > Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com> > > --- > > librdmacm/examples/rdma_server.c | 8 ++++++-- > > 1 file changed, 6 insertions(+), 2 deletions(-) > > > > diff --git a/librdmacm/examples/rdma_server.c b/librdmacm/examples/rdma_server.c > > index bcf649fe..77c021ac 100644 > > --- a/librdmacm/examples/rdma_server.c > > +++ b/librdmacm/examples/rdma_server.c > > @@ -36,6 +36,7 @@ > > #include <rdma/rdma_cma.h> > > #include <rdma/rdma_verbs.h> > > > > +static const char *server = "127.0.0.1"; > > static const char *port = "7471"; > > > > static struct rdma_cm_id *listen_id, *id; > > @@ -55,7 +56,7 @@ static int run(void) > > memset(&hints, 0, sizeof hints); > > hints.ai_flags = RAI_PASSIVE; > > hints.ai_port_space = RDMA_PS_TCP; > > - ret = rdma_getaddrinfo(NULL, port, &hints, &res); > > + ret = rdma_getaddrinfo(server, port, &hints, &res); > > In such case (server is going to be not-NULL), the AI_PASSIVE hint > will be ignored by getaddrinfo, but don't know whenever it is safe to > remove "hints.ai_flags = RAI_PASSIVE;" line above. > > What do you think? Yuval? > > Thanks
On Mon, Mar 05, 2018 at 08:55:25AM +0200, Leon Romanovsky wrote: > On Thu, Mar 01, 2018 at 02:58:46PM +0200, Yuval Shaia wrote: > > Add a command line option to allow selecting specific address as done in > > rdma_client. > > > > Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com> > > --- > > librdmacm/examples/rdma_server.c | 8 ++++++-- > > 1 file changed, 6 insertions(+), 2 deletions(-) > > > > diff --git a/librdmacm/examples/rdma_server.c b/librdmacm/examples/rdma_server.c > > index bcf649fe..77c021ac 100644 > > --- a/librdmacm/examples/rdma_server.c > > +++ b/librdmacm/examples/rdma_server.c > > @@ -36,6 +36,7 @@ > > #include <rdma/rdma_cma.h> > > #include <rdma/rdma_verbs.h> > > > > +static const char *server = "127.0.0.1"; > > static const char *port = "7471"; > > > > static struct rdma_cm_id *listen_id, *id; > > @@ -55,7 +56,7 @@ static int run(void) > > memset(&hints, 0, sizeof hints); > > hints.ai_flags = RAI_PASSIVE; > > hints.ai_port_space = RDMA_PS_TCP; > > - ret = rdma_getaddrinfo(NULL, port, &hints, &res); > > + ret = rdma_getaddrinfo(server, port, &hints, &res); Hi Leon, > > In such case (server is going to be not-NULL), the AI_PASSIVE hint > will be ignored by getaddrinfo, but don't know whenever it is safe to > remove "hints.ai_flags = RAI_PASSIVE;" line above. > > What do you think? Just read man page and realized that you are correct, when 'node' argument is given the AI_PASSIVE is ignored. I am confused, why can't i have them both? i.e be *passive* side of the connection *while* still bind to specific source address. I tried your suggestions, first this flag is used not only in call to getaddrinfo so not sure it is safe to remove the setting. Second, i tried that :) and it fails in rdma_listen phase. > > Thanks -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
> Just read man page and realized that you are correct, when 'node' > argument is given the AI_PASSIVE is ignored. > I am confused, why can't i have them both? i.e be *passive* side of > the connection *while* still bind to specific source address. The behavior matches getaddrinfo(): If the AI_PASSIVE flag is specified in hints.ai_flags, and node is NULL, then the returned socket addresses will be suitable for bind(2)ing a socket that will accept(2) connections. The returned socket address will contain the "wildcard address" (INADDR_ANY for IPv4 addresses, IN6ADDR_ANY_INIT for IPv6 address). The wildcard address is used by applications (typically servers) that intend to accept connections on any of the host's network addresses. If node is not NULL, then the AI_PASSIVE flag is ignored. You can call getaddrinfo() without the passive flag and simply use the returned address as input into bind, provided that you know the address is local. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Mar 12, 2018 at 05:39:07PM +0000, Hefty, Sean wrote: > > Just read man page and realized that you are correct, when 'node' > > argument is given the AI_PASSIVE is ignored. > > I am confused, why can't i have them both? i.e be *passive* side of > > the connection *while* still bind to specific source address. > > The behavior matches getaddrinfo(): > > If the AI_PASSIVE flag is specified in hints.ai_flags, and node is > NULL, then the returned socket addresses will be suitable for > bind(2)ing a socket that will accept(2) connections. The returned > socket address will contain the "wildcard address" (INADDR_ANY for > IPv4 addresses, IN6ADDR_ANY_INIT for IPv6 address). The wildcard > address is used by applications (typically servers) that intend to > accept connections on any of the host's network addresses. If node > is not NULL, then the AI_PASSIVE flag is ignored. > > You can call getaddrinfo() without the passive flag and simply use the returned address as input into bind, provided that you know the address is local. Yeah perfect, it goes hand by hand with my testings. So actually AI_PASSIVE can be skipped in local server case, right? Leon, Since RAI_PASSIVE is used not only for getaddrinfo i would leave it's settings as it is unless you have an objection. What do you think? > -- > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Mar 12, 2018 at 08:40:14PM +0200, Yuval Shaia wrote: > On Mon, Mar 12, 2018 at 05:39:07PM +0000, Hefty, Sean wrote: > > > Just read man page and realized that you are correct, when 'node' > > > argument is given the AI_PASSIVE is ignored. > > > I am confused, why can't i have them both? i.e be *passive* side of > > > the connection *while* still bind to specific source address. > > > > The behavior matches getaddrinfo(): > > > > If the AI_PASSIVE flag is specified in hints.ai_flags, and node is > > NULL, then the returned socket addresses will be suitable for > > bind(2)ing a socket that will accept(2) connections. The returned > > socket address will contain the "wildcard address" (INADDR_ANY for > > IPv4 addresses, IN6ADDR_ANY_INIT for IPv6 address). The wildcard > > address is used by applications (typically servers) that intend to > > accept connections on any of the host's network addresses. If node > > is not NULL, then the AI_PASSIVE flag is ignored. > > > > You can call getaddrinfo() without the passive flag and simply use the returned address as input into bind, provided that you know the address is local. > > Yeah perfect, it goes hand by hand with my testings. So actually AI_PASSIVE > can be skipped in local server case, right? > > Leon, > Since RAI_PASSIVE is used not only for getaddrinfo i would leave it's > settings as it is unless you have an objection. > > What do you think? Thanks, applied. > > > -- > > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > > the body of a message to majordomo@vger.kernel.org > > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Mar 13, 2018 at 06:58:58PM +0200, Leon Romanovsky wrote: > On Mon, Mar 12, 2018 at 08:40:14PM +0200, Yuval Shaia wrote: > > On Mon, Mar 12, 2018 at 05:39:07PM +0000, Hefty, Sean wrote: > > > > Just read man page and realized that you are correct, when 'node' > > > > argument is given the AI_PASSIVE is ignored. > > > > I am confused, why can't i have them both? i.e be *passive* side of > > > > the connection *while* still bind to specific source address. > > > > > > The behavior matches getaddrinfo(): > > > > > > If the AI_PASSIVE flag is specified in hints.ai_flags, and node is > > > NULL, then the returned socket addresses will be suitable for > > > bind(2)ing a socket that will accept(2) connections. The returned > > > socket address will contain the "wildcard address" (INADDR_ANY for > > > IPv4 addresses, IN6ADDR_ANY_INIT for IPv6 address). The wildcard > > > address is used by applications (typically servers) that intend to > > > accept connections on any of the host's network addresses. If node > > > is not NULL, then the AI_PASSIVE flag is ignored. > > > > > > You can call getaddrinfo() without the passive flag and simply use the returned address as input into bind, provided that you know the address is local. > > > > Yeah perfect, it goes hand by hand with my testings. So actually AI_PASSIVE > > can be skipped in local server case, right? > > > > Leon, > > Since RAI_PASSIVE is used not only for getaddrinfo i would leave it's > > settings as it is unless you have an objection. > > > > What do you think? > > Thanks, applied. Thanks, is this means that PR against github repo is not needed? > > > > > > -- > > > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > > > the body of a message to majordomo@vger.kernel.org > > > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, Mar 14, 2018 at 01:00:46PM +0200, Yuval Shaia wrote: > On Tue, Mar 13, 2018 at 06:58:58PM +0200, Leon Romanovsky wrote: > > On Mon, Mar 12, 2018 at 08:40:14PM +0200, Yuval Shaia wrote: > > > On Mon, Mar 12, 2018 at 05:39:07PM +0000, Hefty, Sean wrote: > > > > > Just read man page and realized that you are correct, when 'node' > > > > > argument is given the AI_PASSIVE is ignored. > > > > > I am confused, why can't i have them both? i.e be *passive* side of > > > > > the connection *while* still bind to specific source address. > > > > > > > > The behavior matches getaddrinfo(): > > > > > > > > If the AI_PASSIVE flag is specified in hints.ai_flags, and node is > > > > NULL, then the returned socket addresses will be suitable for > > > > bind(2)ing a socket that will accept(2) connections. The returned > > > > socket address will contain the "wildcard address" (INADDR_ANY for > > > > IPv4 addresses, IN6ADDR_ANY_INIT for IPv6 address). The wildcard > > > > address is used by applications (typically servers) that intend to > > > > accept connections on any of the host's network addresses. If node > > > > is not NULL, then the AI_PASSIVE flag is ignored. > > > > > > > > You can call getaddrinfo() without the passive flag and simply use the returned address as input into bind, provided that you know the address is local. > > > > > > Yeah perfect, it goes hand by hand with my testings. So actually AI_PASSIVE > > > can be skipped in local server case, right? > > > > > > Leon, > > > Since RAI_PASSIVE is used not only for getaddrinfo i would leave it's > > > settings as it is unless you have an objection. > > > > > > What do you think? > > > > Thanks, applied. > > Thanks, is this means that PR against github repo is not needed? Yes, there is no need in PR. > > > > > > > > > > -- > > > > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > > > > the body of a message to majordomo@vger.kernel.org > > > > More majordomo info at http://vger.kernel.org/majordomo-info.html > >
diff --git a/librdmacm/examples/rdma_server.c b/librdmacm/examples/rdma_server.c index bcf649fe..77c021ac 100644 --- a/librdmacm/examples/rdma_server.c +++ b/librdmacm/examples/rdma_server.c @@ -36,6 +36,7 @@ #include <rdma/rdma_cma.h> #include <rdma/rdma_verbs.h> +static const char *server = "127.0.0.1"; static const char *port = "7471"; static struct rdma_cm_id *listen_id, *id; @@ -55,7 +56,7 @@ static int run(void) memset(&hints, 0, sizeof hints); hints.ai_flags = RAI_PASSIVE; hints.ai_port_space = RDMA_PS_TCP; - ret = rdma_getaddrinfo(NULL, port, &hints, &res); + ret = rdma_getaddrinfo(server, port, &hints, &res); if (ret) { printf("rdma_getaddrinfo: %s\n", gai_strerror(ret)); return ret; @@ -163,8 +164,11 @@ int main(int argc, char **argv) { int op, ret; - while ((op = getopt(argc, argv, "p:")) != -1) { + while ((op = getopt(argc, argv, "s:p:")) != -1) { switch (op) { + case 's': + server = optarg; + break; case 'p': port = optarg; break;
Add a command line option to allow selecting specific address as done in rdma_client. Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com> --- librdmacm/examples/rdma_server.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)