card is ejected while we're in this routine.
yamamoto-san's original patch had a small race window for AX88190
chips, which I corrected by limiting the number of iterations we'd try
to reset the bits to be about 15ms rather than forever. This seems to
work for me, but I don't have a large collections of cards based on
this chipset.
Submitted by: YAMAMOTO Shigeru
Overview of problems fixed:
- fix support for saving and restoring filter/NAT state information
(across reboots for example);
- ipmon(8) is started before loading any filter/NAT rules;
- ipmon(8) and ipfs(8) do not solely depend on ipfilter_enable anymore,
they now also work when only ipnat_enable is true;
- the multiple occurrences of code loading the ipfilter kernel module
have been removed;
- the options have been removed from the _program variables in
defaults/rc.conf and the comments in that file have been updated to
reflect (possibly new) reality;
- the rc.conf.5 manual page has been updated to reflect the changes.
Submitted by: Arjan de Vet <devet@devet.org>
PR: conf/25223, kern/25344, conf/25809,
conf/26275, bin/27016, conf/31482
was never technically true (it's snp(4) that required root, not
watch(8)), and after snp.c 1.64, isn't even effectively true, since
who can run watch(8) depends on the permissions of the snp device(s).
Sort options in SYNOPSIS and DESCRIPTION while I'm here.
Previously, watch would always use the first device it could
successfully open, but this isn't always desired. Specifically, it
may not be desired during debugging (of snp), or if a particular snp
device has different permissions (which makes since after snp.c 1.64).
snooped on. This causes all kinds of Bad Things(tm) to happen since
closing one session will clobber state that's needed for the other
one. This could theoretically be supported if the code was careful,
but until somebody implements that, preventing this will stop people
from unknowingly shooting themselves in the foot.
device cloned, and assign all further devices to depend on it. This
allows us to call dev_depends() on it at module unload time to get rid
of /dev/snp* (in the devfs case, anyway). For this to work, we must
not destroy the device at close time. [Idea stolen from if_tun.]
The above has the following sideaffects: (a) The snp device used by
watch(8) will remain after watch(8) exits. This is probably how it
should have been all along, and how it was before devfs came along.
(b) Module unload doesn't panic if there are any /dev/snp* devices
which haven't been used (and thus previously destroyed). Thus, we can
reenable the unload functionality disabled in rev. 1.65.
PR: 32012
to get it all right, allowing ipnat to be enabled independantly of ipfilter
in rc.conf (among other things).
PR: multiple
Submitted by: Arjan de Vet <devet@devet.org>
Reviewed by: Giorgos Keramidas <keramida@FreeBSD.org>
MFC after:3 days
- Add memory barrier definition for sparc64.
Patch sent by David E. O'Brien, approved by maintainer.
- Fix an endianization error of a bus physical address used from SCRIPTS
that made the driver fail on big endian machines as sparc64.
The kernel certainly doesn't use _PATH_DEV or even /dev/ to find the device.
It cannot, since "/" has not been mounted. Maybe the only affect of using
/dev/ is that it gets put in the mounted-from name for "/", so that mount(8),
etc., display an absolute path before "/" has been remounted. Many have
never bothered typing the full path, and code that constructs a path in
rootdevnames[] never bothered to construct a full path, so the example
shouldn't have it.
Submitted by: bde
signature, but otherwise behaves just like a normal USB mass-storage
device. Add a new quirk to cover this case, and enable it for C-1
cameras. The quirk enables translation from the C-1 signature to
the normal CSWSIGNATURE value.
Reviewed by: n_hibma
up in the same way that we expect them to be when we read them.
This is a no-op on i386 and probably on alphas, as we currently
only support AF_INET and AF_INET6.
of 0.0.0.0.
The OpenBSD PF_ROUTE/NET_RT_DUMP sysctl is sending back routes with
RTAX_NETMASK set, but the corresponding sockaddr being 4 zero bytes
(with an address family of zero). ppp was getting confused by this
and ending up interpreting it as a 0.0.0.0/32 routing table
destination and subsequently failing to do anything with the route.
Specifically, after this fix, ppp under OpenBSD can successfully
change and delete the default route again !
ncprange structure.
Don't write() the netmask for IPv6 sockaddrs to the routing socket if
the prefixlen is 128.
It seems that messages written to the routing socket with the scopeid
set for link local addresses are not understood. Instead, we have to
put the scopeid in the 5th and 6th bytes of the address (see
adjust_linklocal() in ncpaddr.c). I think this may be a bug in the
KAME implementation - it should really understand both forms.
to be followed by nfsnodehashtbl, so bzeroing callouts beyond the end of
tcp_syncache soon caused a null pointer panic when nfsnodehashtbl was
accessed.
This file is now generated using src/tools/tools/pciid/mk_pci_vendors.pl,
which merges the Boemler and Heckenbach lists used for rev 1.2.
For now, mk_pci_vendors.pl is called with the -l option, which uses
the entry with the longer description where the same device or vendor
is found in both lists.
If it turns out that this causes to much "back-and-forth" in future
deltas, we can drop the use of the -l option.
