[ovs-dev] [PATCH] utilities: FreeBSD compatibility for date / sha1sum.

Ben Pfaff blp at nicira.com
Tue Jul 31 17:24:12 UTC 2012


On Tue, Jul 31, 2012 at 01:09:02PM -0400, Ed Maste wrote:
> On 31 July 2012 12:51, Ed Maste <emaste at freebsd.org> wrote:
> > With this change (and after enabling SSL in configure) all but 3 skipped
> > tests in "make check" pass on FreeBSD.
> 
> As an aside the three skipped tests are:
> 
> library.at unix domain socket long pathname - I don't think I can do
> anything about this one.

This came up in particular because the Debian autobuilders do their
builds inside deep directory hierarchies, so "make check" was failing
because of the long paths.  Unless FreeBSD has autobuilders that do the
same, it will probably not come up there.

(It's possible that another workaround for long pathnames could be
invented via chdir(), I guess.)

> ofproto.at flow monitoring pause and resume.  FreeBSD doesn't have
> /proc/sys/net/core/rmem_max, but the equivalent should be "sysctl -n
> net.inet.tcp.recvbuf_max."  I need to figure out a little m4 here.

I don't think much m4 is needed here.  We can probably replace the

    # With a Linux kernel, this file has the maximum socket receive buffer
    # size.  That's important for this test, which tests behavior when the
    # receive buffer overflows.
    AT_SKIP_IF([test ! -e /proc/sys/net/core/rmem_max])

    # Calculate the total amount of queuing: rmem_max in the kernel, 128 kB
    # in ofproto sending userspace (see ofmonitor_flush() in connmgr.c).
    rmem_max=`cat /proc/sys/net/core/rmem_max`

by something like this:

if test -e /proc/sys/net/core/rmem_max; then
    # Linux
    rmem_max=`cat /proc/sys/net/core/rmem_max`
elif rmem_max=`sysctl -n net.inet.tcp.recvbuf_max 2>/dev/null`; then
    # FreeBSD.
else
    AT_SKIP_IF([:])
fi

> vlog.at can't reopen log file - currently held up by a bug in
> FreeBSD's sh(1); once that's fixed I expect it to work as long as the
> lindev module is loaded (provides /dev/full).

Hmm, if you have another idea for testing the desired property, that
doesn't require some special compatibility module, I'm open to that too.

Thanks,

Ben.



More information about the dev mailing list