| Message ID | 1493334794-14665-2-git-send-email-gvrose8192@gmail.com |
|---|---|
| State | Rejected |
| Headers | show |
On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: > Parenthetical imbalance was causing some checks to not run > > Signed-off-by: Greg Rose <gvrose8192@gmail.com> > --- > acinclude.m4 | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) > > diff --git a/acinclude.m4 b/acinclude.m4 > index 9f8e30d..0d6647e 100644 > --- a/acinclude.m4 > +++ b/acinclude.m4 > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ > > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) I think this is deliberate.. Only if 'udp_flow_src_port' exists in net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. Otherwise, we'll provide an implementation of udp_flow_src_port. (Note that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET won't be defined, so we'll also backport that function for use from this backported function.
On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: > On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: > > Parenthetical imbalance was causing some checks to not run > > > > Signed-off-by: Greg Rose <gvrose8192@gmail.com> > > --- > > acinclude.m4 | 6 +++--- > > 1 file changed, 3 insertions(+), 3 deletions(-) > > > > diff --git a/acinclude.m4 b/acinclude.m4 > > index 9f8e30d..0d6647e 100644 > > --- a/acinclude.m4 > > +++ b/acinclude.m4 > > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ > > > > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], > > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) > > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], > > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], > > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) > > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) > > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], > > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) > > I think this is deliberate.. Only if 'udp_flow_src_port' exists in > net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define > that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. > Otherwise, we'll provide an implementation of udp_flow_src_port. (Note > that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' > taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET > won't be defined, so we'll also backport that function for use from > this backported function. Right, and now it works so far as I can tell. Before there was an imbalance in the parenthesis and none of the other OVS_GREP_IFELSE clauses following would execute, leaving my own OVS_GREP_IFELSE clause with the ipv6 frag init check not executing at all. Just add my patch to acinclude.m4 without this patch and the clause I added won't ever execute. I guess I could move it above if it is truly intentional that no clauses below that one should ever execute. That's a bit puzzling to me but I could very well be missing something. Thanks, - Greg
On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: > On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: >> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: >> > Parenthetical imbalance was causing some checks to not run >> > >> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> >> > --- >> > acinclude.m4 | 6 +++--- >> > 1 file changed, 3 insertions(+), 3 deletions(-) >> > >> > diff --git a/acinclude.m4 b/acinclude.m4 >> > index 9f8e30d..0d6647e 100644 >> > --- a/acinclude.m4 >> > +++ b/acinclude.m4 >> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ >> > >> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], >> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) >> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], >> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], >> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) >> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) >> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], >> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) >> >> I think this is deliberate.. Only if 'udp_flow_src_port' exists in >> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define >> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. >> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note >> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' >> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET >> won't be defined, so we'll also backport that function for use from >> this backported function. > > Right, and now it works so far as I can tell. Before there was an > imbalance in the parenthesis and none of the other OVS_GREP_IFELSE > clauses following would execute, leaving my own OVS_GREP_IFELSE clause > with the ipv6 frag init check not executing at all. > > Just add my patch to acinclude.m4 without this patch and the clause I > added won't ever execute. I guess I could move it above if it is truly > intentional that no clauses below that one should ever execute. I wonder if the open bracket in the "inet_get_local_port_range(net" is somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? I know my editor gets a bit confused by this one, but I assumed that it's just my editor rather than the actual execution. When I ran configure without this patch on my trusty (kernel-3.13) machine, I saw the following line: checking whether nf_defrag_ipv6_enable has parameter net in /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... no
On 28 April 2017 at 14:18, Joe Stringer <joe@ovn.org> wrote: > On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: >> On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: >>> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: >>> > Parenthetical imbalance was causing some checks to not run >>> > >>> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> >>> > --- >>> > acinclude.m4 | 6 +++--- >>> > 1 file changed, 3 insertions(+), 3 deletions(-) >>> > >>> > diff --git a/acinclude.m4 b/acinclude.m4 >>> > index 9f8e30d..0d6647e 100644 >>> > --- a/acinclude.m4 >>> > +++ b/acinclude.m4 >>> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ >>> > >>> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], >>> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) >>> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], >>> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], >>> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], >>> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) >>> >>> I think this is deliberate.. Only if 'udp_flow_src_port' exists in >>> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define >>> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. >>> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note >>> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' >>> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET >>> won't be defined, so we'll also backport that function for use from >>> this backported function. >> >> Right, and now it works so far as I can tell. Before there was an >> imbalance in the parenthesis and none of the other OVS_GREP_IFELSE >> clauses following would execute, leaving my own OVS_GREP_IFELSE clause >> with the ipv6 frag init check not executing at all. >> >> Just add my patch to acinclude.m4 without this patch and the clause I >> added won't ever execute. I guess I could move it above if it is truly >> intentional that no clauses below that one should ever execute. > > I wonder if the open bracket in the "inet_get_local_port_range(net" is > somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? > > I know my editor gets a bit confused by this one, but I assumed that > it's just my editor rather than the actual execution. > > When I ran configure without this patch on my trusty (kernel-3.13) > machine, I saw the following line: > > checking whether nf_defrag_ipv6_enable has parameter net in > /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... > no ^ I should mean to say, with all of the other patches in this series I see this.
On Fri, 2017-04-28 at 14:18 -0700, Joe Stringer wrote: > On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: > > On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: > >> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: > >> > Parenthetical imbalance was causing some checks to not run > >> > > >> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> > >> > --- > >> > acinclude.m4 | 6 +++--- > >> > 1 file changed, 3 insertions(+), 3 deletions(-) > >> > > >> > diff --git a/acinclude.m4 b/acinclude.m4 > >> > index 9f8e30d..0d6647e 100644 > >> > --- a/acinclude.m4 > >> > +++ b/acinclude.m4 > >> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ > >> > > >> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], > >> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) > >> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], > >> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], > >> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) > >> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) > >> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], > >> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) > >> > >> I think this is deliberate.. Only if 'udp_flow_src_port' exists in > >> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define > >> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. > >> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note > >> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' > >> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET > >> won't be defined, so we'll also backport that function for use from > >> this backported function. > > > > Right, and now it works so far as I can tell. Before there was an > > imbalance in the parenthesis and none of the other OVS_GREP_IFELSE > > clauses following would execute, leaving my own OVS_GREP_IFELSE clause > > with the ipv6 frag init check not executing at all. > > > > Just add my patch to acinclude.m4 without this patch and the clause I > > added won't ever execute. I guess I could move it above if it is truly > > intentional that no clauses below that one should ever execute. > > I wonder if the open bracket in the "inet_get_local_port_range(net" is > somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? > > I know my editor gets a bit confused by this one, but I assumed that > it's just my editor rather than the actual execution. > > When I ran configure without this patch on my trusty (kernel-3.13) > machine, I saw the following line: > > checking whether nf_defrag_ipv6_enable has parameter net in > /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... > no Really? Let me try it again. I'm running on 4.10.12 to test the bug. Thanks, - Greg
On Fri, 2017-04-28 at 14:21 -0700, Joe Stringer wrote: > On 28 April 2017 at 14:18, Joe Stringer <joe@ovn.org> wrote: > > On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: > >> On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: > >>> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: > >>> > Parenthetical imbalance was causing some checks to not run > >>> > > >>> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> > >>> > --- > >>> > acinclude.m4 | 6 +++--- > >>> > 1 file changed, 3 insertions(+), 3 deletions(-) > >>> > > >>> > diff --git a/acinclude.m4 b/acinclude.m4 > >>> > index 9f8e30d..0d6647e 100644 > >>> > --- a/acinclude.m4 > >>> > +++ b/acinclude.m4 > >>> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ > >>> > > >>> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], > >>> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) > >>> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], > >>> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], > >>> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) > >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) > >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], > >>> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) > >>> > >>> I think this is deliberate.. Only if 'udp_flow_src_port' exists in > >>> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define > >>> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. > >>> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note > >>> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' > >>> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET > >>> won't be defined, so we'll also backport that function for use from > >>> this backported function. > >> > >> Right, and now it works so far as I can tell. Before there was an > >> imbalance in the parenthesis and none of the other OVS_GREP_IFELSE > >> clauses following would execute, leaving my own OVS_GREP_IFELSE clause > >> with the ipv6 frag init check not executing at all. > >> > >> Just add my patch to acinclude.m4 without this patch and the clause I > >> added won't ever execute. I guess I could move it above if it is truly > >> intentional that no clauses below that one should ever execute. > > > > I wonder if the open bracket in the "inet_get_local_port_range(net" is > > somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? > > > > I know my editor gets a bit confused by this one, but I assumed that > > it's just my editor rather than the actual execution. > > > > When I ran configure without this patch on my trusty (kernel-3.13) > > machine, I saw the following line: > > > > checking whether nf_defrag_ipv6_enable has parameter net in > > /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... > > no > > ^ I should mean to say, with all of the other patches in this series I see this. Well I'll be a monkey's uncle, you're right. I reverted this patch in my own experimental tree and now it works. checking whether nf_defrag_ipv6_enable has parameter net in /lib/modules/4.10.12/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... yes But it wasn't before, I swear! LOL And none of the other ones following were printing either. None of these, but now they are. checking whether ignore_df matches in /lib/modules/4.10.12/build/include/linux/skbuff.h... yes checking whether NET_NAME_UNKNOWN matches in /lib/modules/4.10.12/build/include/uapi/linux/netdevice.h... yes checking whether sk_no_check_tx matches in /lib/modules/4.10.12/build/include/net/sock.h... yes checking whether no_check6_tx matches in /lib/modules/4.10.12/build/include/linux/udp.h... yes checking whether el6 matches in /lib/modules/4.10.12/build/include/linux/utsrelease.h... file not found checking whether udp_add_offload has parameter net in /lib/modules/4.10.12/build/include/net/protocol.h... no In any case, I can't repro now so I'll respin the series without this patch and send it along. SMH Thanks Joe! - Greg
On 28 April 2017 at 14:31, Greg Rose <gvrose8192@gmail.com> wrote: > On Fri, 2017-04-28 at 14:21 -0700, Joe Stringer wrote: >> On 28 April 2017 at 14:18, Joe Stringer <joe@ovn.org> wrote: >> > On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: >> >> On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: >> >>> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: >> >>> > Parenthetical imbalance was causing some checks to not run >> >>> > >> >>> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> >> >>> > --- >> >>> > acinclude.m4 | 6 +++--- >> >>> > 1 file changed, 3 insertions(+), 3 deletions(-) >> >>> > >> >>> > diff --git a/acinclude.m4 b/acinclude.m4 >> >>> > index 9f8e30d..0d6647e 100644 >> >>> > --- a/acinclude.m4 >> >>> > +++ b/acinclude.m4 >> >>> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ >> >>> > >> >>> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], >> >>> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) >> >>> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], >> >>> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], >> >>> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) >> >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) >> >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], >> >>> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) >> >>> >> >>> I think this is deliberate.. Only if 'udp_flow_src_port' exists in >> >>> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define >> >>> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. >> >>> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note >> >>> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' >> >>> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET >> >>> won't be defined, so we'll also backport that function for use from >> >>> this backported function. >> >> >> >> Right, and now it works so far as I can tell. Before there was an >> >> imbalance in the parenthesis and none of the other OVS_GREP_IFELSE >> >> clauses following would execute, leaving my own OVS_GREP_IFELSE clause >> >> with the ipv6 frag init check not executing at all. >> >> >> >> Just add my patch to acinclude.m4 without this patch and the clause I >> >> added won't ever execute. I guess I could move it above if it is truly >> >> intentional that no clauses below that one should ever execute. >> > >> > I wonder if the open bracket in the "inet_get_local_port_range(net" is >> > somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? >> > >> > I know my editor gets a bit confused by this one, but I assumed that >> > it's just my editor rather than the actual execution. >> > >> > When I ran configure without this patch on my trusty (kernel-3.13) >> > machine, I saw the following line: >> > >> > checking whether nf_defrag_ipv6_enable has parameter net in >> > /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... >> > no >> >> ^ I should mean to say, with all of the other patches in this series I see this. > > Well I'll be a monkey's uncle, you're right. I reverted this patch in > my own experimental tree and now it works. > > checking whether nf_defrag_ipv6_enable has parameter net > in /lib/modules/4.10.12/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... yes > > But it wasn't before, I swear! LOL And none of the other ones > following were printing either. None of these, but now they are. > > checking whether ignore_df matches > in /lib/modules/4.10.12/build/include/linux/skbuff.h... yes > checking whether NET_NAME_UNKNOWN matches > in /lib/modules/4.10.12/build/include/uapi/linux/netdevice.h... yes > checking whether sk_no_check_tx matches > in /lib/modules/4.10.12/build/include/net/sock.h... yes > checking whether no_check6_tx matches > in /lib/modules/4.10.12/build/include/linux/udp.h... yes > checking whether el6 matches > in /lib/modules/4.10.12/build/include/linux/utsrelease.h... file not > found > checking whether udp_add_offload has parameter net > in /lib/modules/4.10.12/build/include/net/protocol.h... no > > In any case, I can't repro now so I'll respin the series without this > patch and send it along. SMH Hmm. I wonder if you needed to run boot.sh again or something to regenerate the configure script? Anyway, good to know. Don't worry about respinning, I can just apply the patches.
On Fri, 2017-04-28 at 14:33 -0700, Joe Stringer wrote: > On 28 April 2017 at 14:31, Greg Rose <gvrose8192@gmail.com> wrote: > > On Fri, 2017-04-28 at 14:21 -0700, Joe Stringer wrote: > >> On 28 April 2017 at 14:18, Joe Stringer <joe@ovn.org> wrote: > >> > On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: > >> >> On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: > >> >>> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: > >> >>> > Parenthetical imbalance was causing some checks to not run > >> >>> > > >> >>> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> > >> >>> > --- > >> >>> > acinclude.m4 | 6 +++--- > >> >>> > 1 file changed, 3 insertions(+), 3 deletions(-) > >> >>> > > >> >>> > diff --git a/acinclude.m4 b/acinclude.m4 > >> >>> > index 9f8e30d..0d6647e 100644 > >> >>> > --- a/acinclude.m4 > >> >>> > +++ b/acinclude.m4 > >> >>> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ > >> >>> > > >> >>> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], > >> >>> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) > >> >>> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], > >> >>> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], > >> >>> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) > >> >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) > >> >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], > >> >>> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) > >> >>> > >> >>> I think this is deliberate.. Only if 'udp_flow_src_port' exists in > >> >>> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define > >> >>> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. > >> >>> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note > >> >>> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' > >> >>> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET > >> >>> won't be defined, so we'll also backport that function for use from > >> >>> this backported function. > >> >> > >> >> Right, and now it works so far as I can tell. Before there was an > >> >> imbalance in the parenthesis and none of the other OVS_GREP_IFELSE > >> >> clauses following would execute, leaving my own OVS_GREP_IFELSE clause > >> >> with the ipv6 frag init check not executing at all. > >> >> > >> >> Just add my patch to acinclude.m4 without this patch and the clause I > >> >> added won't ever execute. I guess I could move it above if it is truly > >> >> intentional that no clauses below that one should ever execute. > >> > > >> > I wonder if the open bracket in the "inet_get_local_port_range(net" is > >> > somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? > >> > > >> > I know my editor gets a bit confused by this one, but I assumed that > >> > it's just my editor rather than the actual execution. > >> > > >> > When I ran configure without this patch on my trusty (kernel-3.13) > >> > machine, I saw the following line: > >> > > >> > checking whether nf_defrag_ipv6_enable has parameter net in > >> > /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... > >> > no > >> > >> ^ I should mean to say, with all of the other patches in this series I see this. > > > > Well I'll be a monkey's uncle, you're right. I reverted this patch in > > my own experimental tree and now it works. > > > > checking whether nf_defrag_ipv6_enable has parameter net > > in /lib/modules/4.10.12/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... yes > > > > But it wasn't before, I swear! LOL And none of the other ones > > following were printing either. None of these, but now they are. > > > > checking whether ignore_df matches > > in /lib/modules/4.10.12/build/include/linux/skbuff.h... yes > > checking whether NET_NAME_UNKNOWN matches > > in /lib/modules/4.10.12/build/include/uapi/linux/netdevice.h... yes > > checking whether sk_no_check_tx matches > > in /lib/modules/4.10.12/build/include/net/sock.h... yes > > checking whether no_check6_tx matches > > in /lib/modules/4.10.12/build/include/linux/udp.h... yes > > checking whether el6 matches > > in /lib/modules/4.10.12/build/include/linux/utsrelease.h... file not > > found > > checking whether udp_add_offload has parameter net > > in /lib/modules/4.10.12/build/include/net/protocol.h... no > > > > In any case, I can't repro now so I'll respin the series without this > > patch and send it along. SMH > > Hmm. I wonder if you needed to run boot.sh again or something to > regenerate the configure script? Right. Could be, I've been playing around with the different types of 'clean' targets and have finally sort of settled on the maintainer-clean-am now. Seems to give the most reliable results between different kernels and tool chains. Unless there's a better one? > > Anyway, good to know. Don't worry about respinning, I can just apply > the patches. TYVM!
On 28 April 2017 at 14:45, Greg Rose <gvrose8192@gmail.com> wrote: > On Fri, 2017-04-28 at 14:33 -0700, Joe Stringer wrote: >> On 28 April 2017 at 14:31, Greg Rose <gvrose8192@gmail.com> wrote: >> > On Fri, 2017-04-28 at 14:21 -0700, Joe Stringer wrote: >> >> On 28 April 2017 at 14:18, Joe Stringer <joe@ovn.org> wrote: >> >> > On 28 April 2017 at 14:02, Greg Rose <gvrose8192@gmail.com> wrote: >> >> >> On Fri, 2017-04-28 at 13:55 -0700, Joe Stringer wrote: >> >> >>> On 27 April 2017 at 16:13, Greg Rose <gvrose8192@gmail.com> wrote: >> >> >>> > Parenthetical imbalance was causing some checks to not run >> >> >>> > >> >> >>> > Signed-off-by: Greg Rose <gvrose8192@gmail.com> >> >> >>> > --- >> >> >>> > acinclude.m4 | 6 +++--- >> >> >>> > 1 file changed, 3 insertions(+), 3 deletions(-) >> >> >>> > >> >> >>> > diff --git a/acinclude.m4 b/acinclude.m4 >> >> >>> > index 9f8e30d..0d6647e 100644 >> >> >>> > --- a/acinclude.m4 >> >> >>> > +++ b/acinclude.m4 >> >> >>> > @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ >> >> >>> > >> >> >>> > OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], >> >> >>> > [OVS_DEFINE([HAVE_VXLAN_METADATA])]) >> >> >>> > - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], >> >> >>> > - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], >> >> >>> > - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) >> >> >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) >> >> >>> > + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], >> >> >>> > + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) >> >> >>> >> >> >>> I think this is deliberate.. Only if 'udp_flow_src_port' exists in >> >> >>> net/udp.h, *AND* 'inet_get_local_port_range(net' exists, we define >> >> >>> that HAVE_UDP_FLOW_SRC_PORT exists with the correct implementation. >> >> >>> Otherwise, we'll provide an implementation of udp_flow_src_port. (Note >> >> >>> that if 'udp_flow_src_port' exists but 'inet_get_local_port_range' >> >> >>> taking 'net' does not, then HAVE_INET_GET_LOCAL_PORT_RANGE_USING_NET >> >> >>> won't be defined, so we'll also backport that function for use from >> >> >>> this backported function. >> >> >> >> >> >> Right, and now it works so far as I can tell. Before there was an >> >> >> imbalance in the parenthesis and none of the other OVS_GREP_IFELSE >> >> >> clauses following would execute, leaving my own OVS_GREP_IFELSE clause >> >> >> with the ipv6 frag init check not executing at all. >> >> >> >> >> >> Just add my patch to acinclude.m4 without this patch and the clause I >> >> >> added won't ever execute. I guess I could move it above if it is truly >> >> >> intentional that no clauses below that one should ever execute. >> >> > >> >> > I wonder if the open bracket in the "inet_get_local_port_range(net" is >> >> > somehow interfering with the bracket balancing in the OVS_GREP_IFELSE? >> >> > >> >> > I know my editor gets a bit confused by this one, but I assumed that >> >> > it's just my editor rather than the actual execution. >> >> > >> >> > When I ran configure without this patch on my trusty (kernel-3.13) >> >> > machine, I saw the following line: >> >> > >> >> > checking whether nf_defrag_ipv6_enable has parameter net in >> >> > /lib/modules/3.13.0-91-generic/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... >> >> > no >> >> >> >> ^ I should mean to say, with all of the other patches in this series I see this. >> > >> > Well I'll be a monkey's uncle, you're right. I reverted this patch in >> > my own experimental tree and now it works. >> > >> > checking whether nf_defrag_ipv6_enable has parameter net >> > in /lib/modules/4.10.12/build/include/net/netfilter/ipv6/nf_defrag_ipv6.h... yes >> > >> > But it wasn't before, I swear! LOL And none of the other ones >> > following were printing either. None of these, but now they are. >> > >> > checking whether ignore_df matches >> > in /lib/modules/4.10.12/build/include/linux/skbuff.h... yes >> > checking whether NET_NAME_UNKNOWN matches >> > in /lib/modules/4.10.12/build/include/uapi/linux/netdevice.h... yes >> > checking whether sk_no_check_tx matches >> > in /lib/modules/4.10.12/build/include/net/sock.h... yes >> > checking whether no_check6_tx matches >> > in /lib/modules/4.10.12/build/include/linux/udp.h... yes >> > checking whether el6 matches >> > in /lib/modules/4.10.12/build/include/linux/utsrelease.h... file not >> > found >> > checking whether udp_add_offload has parameter net >> > in /lib/modules/4.10.12/build/include/net/protocol.h... no >> > >> > In any case, I can't repro now so I'll respin the series without this >> > patch and send it along. SMH >> >> Hmm. I wonder if you needed to run boot.sh again or something to >> regenerate the configure script? > > Right. Could be, I've been playing around with the different types of > 'clean' targets and have finally sort of settled on the > maintainer-clean-am now. Seems to give the most reliable results > between different kernels and tool chains. > > Unless there's a better one? I've just started using "git clean" for this purpose, not sure if it's more or less reliable than your methods. It will clobber any untracked files though, so don't save any handy scripts in the directory :-)
On Fri, 2017-04-28 at 14:47 -0700, Joe Stringer wrote: [snip] > > I've just started using "git clean" for this purpose, not sure if it's > more or less reliable than your methods. It will clobber any untracked > files though, so don't save any handy scripts in the directory :-) Ah yes, the nuclear option. 8^)
diff --git a/acinclude.m4 b/acinclude.m4 index 9f8e30d..0d6647e 100644 --- a/acinclude.m4 +++ b/acinclude.m4 @@ -681,9 +681,9 @@ AC_DEFUN([OVS_CHECK_LINUX_COMPAT], [ OVS_GREP_IFELSE([$KSRC/include/net/vxlan.h], [struct vxlan_metadata], [OVS_DEFINE([HAVE_VXLAN_METADATA])]) - OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port], - [OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net], - [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])])]) + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_flow_src_port]) + OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [inet_get_local_port_range(net)], + [OVS_DEFINE([HAVE_UDP_FLOW_SRC_PORT])]) OVS_GREP_IFELSE([$KSRC/include/net/udp.h], [udp_v4_check]) OVS_GREP_IFELSE([$KSRC/include/net/udp_tunnel.h], [udp_tunnel_gro_complete]) OVS_GREP_IFELSE([$KSRC/include/net/udp_tunnel.h], [sk_buff.*udp_tunnel_handle_offloads],
Parenthetical imbalance was causing some checks to not run Signed-off-by: Greg Rose <gvrose8192@gmail.com> --- acinclude.m4 | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)