I noticed your patch adding DESTDIR support in the latest iproute2 release.
Much appreciated! Soon the debian packages might be able to move to actually
using "make install" rather then it's own installation procedure when
building packages. I've noticed something that will break though....
Debian packages usually sets DESTDIR=debian/tmp/ and packages the contents
of that directory as if it where the root file system. This will break
the /usr/lib/{tc,ip}/ module loading, because they DESTDIR (/usr) will be
/whatever-the-build-path-was/debian/tmp/lib/{tc,ip}/.
I beleive others usually call this the LIBDIR to make the separation between
DISTDIR being the (possibly temporary) place things are put when build is
done, and LIBDIR (and others) are used for actual runtime paths.
I'm attaching a patch that I think fixes this, but would be really happy if
you could have a look at to verify I'm not screwing something up.
--
Regards,
Andreas Henriksson
Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
ip a flush: unnecessarily chatty when there's nothing to flush
On tis, 2008-07-22 at 20:21 +0200, martin f krafft wrote:
> piper:~|master|% sudo ip a flush dev eth0 >/dev/null
> Nothing to flush.
>
> It should just shut up, shouldn't it? :)
The patch below makes the "Nothing to flush" only visible when show
statistics flag is given, and then only to stdout rather then stderr.
$ sudo ./ip/ip -s addr flush dev skif
Nothing to flush.
$ sudo ./ip/ip addr flush dev skif
$
Stephen Hemminger [Thu, 22 May 2008 20:41:40 +0000 (13:41 -0700)]
> Use the standard POSIX inet_pton to convert from string to IPV4
> address. This avoids problems where ip parses "127.2" wrong.
Apparently inet_pton doesn't support abbreviated/shortened/classful
ipv4 addresses at all, but inet_aton does.
Since the function only deals with AF_INET anyway maybe using
inet_aton "to increse backwards compatability" (please those
who still want to use the format) could be considered?
(This will still not restore the 10/8 format which apparently used
to work in iproute, so people would have to settle for 10.0/8)
Rafael Almeida [Mon, 2 Jun 2008 00:33:44 +0000 (21:33 -0300)]
Fixed installation when changing DESTDIR
After changing the DESTDIR the installated binaries have some issues
due to hard coded paths. For example, using distributions on NetEm
would segfault.
I've changed iplink.c and tc_util.c so they are now aware of DESTDIR.
Along with that change I needed to change the main Makefile so it
defines the DESTDIR macro when calling gcc.
I also changed the paths so that during the installation sbin, etc,
share and lib directories are created directly inside of the DESTDIR,
instead of creating a usr directory inside that. That's the behaviour
of most packages out there, so I think most users will be expecting
that to happen.
The command "ip addrlabel add/del" displays incorrect error message when provided with insufficient inputs. This patch fixes it in par with "ip addr add/del".
Currently:
# ./ip addrlabel add
RTNETLINK answers: Numerical result out of range
# ./ip addr add
Not enough information: "dev" argument is required.
After patch:
# ./ip addrlabel add
Not enough information: "prefix" argument is required.
Signed-off-by: Varun Chandramohan <varunc@linux.vnet.ibm.com> Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
The current kernel generates 71 possible header fields, but
MAX_FIELDS in lnstat is only 64. This leads to referencing outside
of the array. To fix, increase size of array and chop off parsing
at MAX_FIELDS - 1.
Bertrand Jacquin [Sat, 21 Jun 2008 05:51:19 +0000 (01:51 -0400)]
netem: fix cross-compiling failure
The programs in netem are compiled and run on the build machine, but they
use the CFLAGS that are meant for the target system and often times, these
are incompatible.
Benjamin Thery [Fri, 20 Jun 2008 09:07:35 +0000 (11:07 +0200)]
iproute2: add support for IFLA_NET_NS_PID in ip link
Hi Stephen,
I resend you this patch once more. This time I updated the documentation
too (may be that was the reason why you didn't take it before?).
Please tell me if there are other things missing in this patch
It applies on iproute2 git tree.
Regards,
Benjamin
Description:
------------
This patch adds support for the IFLA_NET_NS_PID type. It is used to
move network devices between network namespaces.
The syntax is:
ip link set DEVICE netns PID
PID is the pid of a process in the target network namespace.
(Daniel Lezcano is the original author).
Signed-off-by: Daniel Lezcano <dlezcano@fr.ibm.com> Signed-off-by: Benjamin Thery <benjamin.thery@bull.net> Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
Patrick McHardy [Mon, 23 Jun 2008 13:59:18 +0000 (15:59 +0200)]
tc: don't set protococol field on filter delete
> # tc filter show dev eth1 | grep 4:29:d1
> filter parent 1: protocol ip pref 5 u32 fh 4:29:d1 order 209 key ht 4
> bkt 29 flowid 1:b7aa
>
> # tc filter del dev eth1 parent 1: pref 5 handle 4:29:d1 u32
> RTNETLINK answers: Invalid argument
> We have an error talking to the kernel
>
> after rollback to package"sys-apps/iproute2-2.6.24.20080108" all
> deleted normal...
The current iproute version uses "protocol all" by default
if its not specified. This is actually only useful for creating
new filters, on deletion an unset protocol is treated as wildcard.
The function get_hz() returns the psched hz value which is wrong
for anything other than tc usage. Should be user hz instead, but
kernel is broken (patch sent) and this code doesn't get hit on
current systems (netlink is used first).
Herbert Xu [Wed, 23 Apr 2008 07:42:32 +0000 (15:42 +0800)]
xfrm: Allow replay setting
Hi Stephen:
[IP] xfrm: Allow replay setting
For certain applications there is a requirement to start the
sequence number from a point other than the default. As it
is the kernel provides an interface to do that but it isn't
available through the ip(8) command. Since we're encouraging
people to migrate over to ip(8) for manual keying, it is useful
to have this ability there.
This patch adds support for setting replay sequence numbers
through ip(8).
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Thanks,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
--
Don't break scripts that depend on previous offset/value format.
Introduce a new -pretty flag for decoding, and (*gasp*) document
the formatting arguments.
Write something about the tweak to enable promoting secondary addresses
instead of deleting them together with the primary address as discussed
in this thread on the netdev mailing list:
http://www.spinics.net/lists/netdev/msg52294.html
The claim that this is supported since 2.6.15 is based on looking at
changes to net/ipv4/devinet.c in the linux-2.6 git tree:
I was asked to at least mention the xfrm option in ip manual. I added
all usage into ip.8 and try to write some basic information about xfrm.
If someone want complete it, I'll be happy.
In ss.c, generic_proc_open(), for which the net_*_open functions are just
convenient wrappers, uses fopen, so errors are signalled by a NULL return
value. Some checks were expecting negative values instead, fix them.
Signed-off-by: Björn Steinbrink <B.Steinbrink@gmx.de> Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
Add more aliases to synchronize IPv4 and IPv6 tunnel command, e.g.,
IPv4: hoplimit (alias to ttl), tclass (alias to tos)
IPv6: dsfield, tos (alias to tc, or tclass), ttl (alias to hoplimit)
Signed-off-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org> Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
The patches branch of the debian packaging repo, at
git://git.debian.org/git/collab-maint/pkg-iproute, now has four patches still
pending. (Not resending, since they've all been posted multiple times before.
Pick them up from the repo if you're interested.)
PJ Waskiewicz [Wed, 13 Feb 2008 11:49:09 +0000 (03:49 -0800)]
Update various classifiers' help output for expected CLASSID syntax
update: Fix the spelling of "hexidecimal"
This updates the help output to specify that CLASSID should be hexidecimal.
This makes sure that a user entering "flowid 1:10" gets his flow put into
band 15 (0x10) and knows why.
Signed-off-by: Peter P Waskiewicz Jr <peter.p.waskiewicz.jr@intel.com> Signed-off-by: Stephen Hemminger <stephen.hemminger@vyatta.com>
This fixes the problem where a bulk operation (like ip flush)
is performed as non-root user. The kernel will only send a response
if there is an error, so check for it.
After further investigation it seems clear to me that reverting the
commit 720a2e8d990707749b2... is the correct thing to do, since the real
fix for the problem this commit was supposed to fix was instead fixed in
commit c29391c7c68f031e246c...
Whatever you specify after a u32 police you will now get a syntax error,
and according to "tc filter add u32 help" there are several things that
you are supposed to be able to specify after a police.
[...]
> Commands like "tc filter add dev ppp0 parent ffff: protocol ip prio 50
> u32 match ip src 0.0.0.0/0 police rate 4mbit burst 10k drop flowid :1"
> apparently no longer works. The flowid is not accepted anymore.
> Reverting commit 720a2e8d99... which you authored seems to "fix" this.
[...]
After further investigation it seems clear to me that reverting the
commit 720a2e8d990707749b2... is the correct thing to do, since the real
fix for the problem this commit was supposed to fix was instead fixed in
commit c29391c7c68f031e246c...
Whatever you specify after a u32 police you will now get a syntax error,
and according to "tc filter add u32 help" there are several things that
you are supposed to be able to specify after a police.
projects / lisovros / iproute2_canprio.git / log