{image}+= LAST_MINUTE README.files
> .endfor
I think that those files they used to be present in iso images, and I
am sure that it would be better if thse files were present. Instead of
commenting out this code, please find a way to include the files.
--apb (Alan Barrett)
ated
> on the iso-image target.
Yes, that was essentially my point. If there's a bug such that those
files don't exist in RELEASEDIR, then fix that bug, don't simply remove
the files from the iso-image creation process.
--apb (Alan Barrett)
tand why you need gmake to
build iso-image for Macs.
--apb (Alan Barrett)
n improve official build.sh itself,
> but it is separate task from distrib/cdrom work for me.
I still think it's a bug that "build.sh release" doesn't store those
files in RELEASEDIR. I may look at fixing it, but I don't have much
time for that now.
--apb (Alan Barrett)
cat ${METALOG} | ${GREP} -v " optional" > ${.TARGET}
+ cat ${METALOG} | ${GREP} -v " optional" | sort -u > ${.TARGET}
.endif
.if defined(DESTDIR) && ${DESTDIR} != ""
It would also make sense to add a sort option to mtree.
--apb (Alan Barrett)
On Fri, 03 Apr 2009, Perry E. Metzger wrote:
> Alan Barrett writes:
> > It would also make sense to add a sort option to mtree.
>
> Does mtree -C sort already? I haven't read the sources but it seems to
> rearrange what it is fed into a canonicalized order.
No, it
to mtree.
> Or fix join.awk, but it's another future work.
>
> My change is just an interim fix for actually forthcoming 5.0.
I have committed a change to make "mtree -C" sort its output. I think
we can now remove sort incovations from shell scripts or make rules that
invoke mtree -C.
--apb (Alan Barrett)
On Fri, 03 Apr 2009, Soren Jacobsen wrote:
> On 04/04 00:47, Alan Barrett wrote:
> > I have committed a change to make "mtree -C" sort its output. I think
> > we can now remove sort incovations from shell scripts or make rules that
> > invoke mtree -C.
>
>
Perry, please could you answer this question.
--apb (Alan Barrett)
On Fri, 13 Mar 2009, Alan Barrett wrote:
> On Fri, 13 Mar 2009, Perry E. Metzger wrote:
> > As indicated, I see nothing wrong with a MKDETERMINISTIC that
> > flips the MKARZERO flag as well as the other fla
ssibly
named "_MKARZERO"), which would automatically be set from a user-tunable
variable such as MKDETERMINISTIC or MKREPRODUCIBLE.
--apb (Alan Barrett)
On Sat, 04 Apr 2009, Alan Barrett wrote:
> Modified Files:
> src/usr.sbin/mtree: mtree.h spec.c
>
> Log Message:
> Tags are internally stored with leading and trailing commas, for ease
> of using strcmp(3) to check them against inclusion or exclusion lists.
> Don&#x
On Tue, 07 Apr 2009, Alan Barrett wrote:
> Modified Files:
> src/usr.bin/xinstall: xinstall.c
>
> Log Message:
> Changes for installing with a metalog:
> * When installing hard links and using a metalog, if -o, -g, -m, or -f
> args were explicitly specified on t
-#defineVC_XLOCK0x8000
> -#defineVC_MASK 0x7fff
> +#defineVC_XLOCK0x4000
> +#defineVC_MASK (0x & ~VC_XLOCK)
It may be easier and safer to use 0x7fffUL.
--apb (Alan Barrett)
On Mon, 18 May 2009, Thomas Klausner wrote:
> Modified Files:
> src/sbin/init: init.8
>
> Log Message:
> Sort SEE ALSO.
It was already sorted. If the sort order is supposed to be
case-insensitive, then please update the instructions in groff_mdoc(7).
--apb (Alan Barrett)
On Tue, 19 May 2009, Thomas Klausner wrote:
> On Tue, May 19, 2009 at 11:40:10AM +0300, Alan Barrett wrote:
> > It was already sorted. If the sort order is supposed to be
> > case-insensitive, then please update the instructions in groff_mdoc(7).
>
> It was just saying
that change in src/tools/compat, I think you should need only the
following lines here:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
--apb (Alan Barrett)
that imply that the British spelling was
incorrect. (Similarly for "supercede"/"supersede".)
--apb (Alan Barrett)
he fs_vfstype column, or simply
remove this test and rely on the test for "0" in the fs_passno column?
--apb (Alan Barrett)
inclined to do the
addition using an unsigned type, which has well-defined overflow
behaviour, and then check the range of the unsigned result.
--apb (Alan Barrett)
aired with another single quote mark. These restrictions apply even
inside an #if 0 or #ifdef notyet block, although many compilers don't
enforce that.
I don't feel competent to delete the entire block, so if that's what you
want then please do it yourself. I would be happy to an an #error if
that's acceptable to you.
--apb (Alan Barrett)
ultiple concurrent invocations of install(1)
> simultaneously install files in the same directory. Fixes bin/41512.
The comment about /* This usage is safe */ for the mktemp() call
is cleary inaccurate. I think this should be fixed by using mkstemp().
--apb (Alan Barrett)
h lines longer than 80 chars.
> Ok supremeleader@
Who?
--apb (Alan Barrett)
. My changes include a
comment block explaining the weird and inconsistent encoding scheme.
See <ftp://ftp.netbsd.org/pub/NetBSD/misc/apb/sort.20070822.diff>.
--apb (Alan Barrett)
On Sun, 16 Aug 2009, David Young wrote:
> > > net.interfaces.3.name=xvif0.0
> > > net.interfaces.3.sndq.maxlen=...
> >
> > sure, that would work.
>
> Except for the fact that a sysctl node name cannot begin with a digit.
What's stopping us from removing that restriction?
--apb (Alan Barrett)
ode, when a simple flushroutes=NO in
/etc/rc.conf will do the job? Perhaps sysinst should automatically set
flushroutes=NO if it knows you have root on NFS.
--apb (Alan Barrett)
a
test for whether to flush routes or to change the default route. For
example, if root is local but /var is on NFS, you will get similar
problems. Perhaps a better test would be that dhcpcd shouldn't touch
the default route unless the default route is through the interface that
dhcpcd is managing.
--apb (Alan Barrett)
ver is in a different subnet.
>
> This change should be pulled up to netbsd-5.
No, this change should be backed out, and the real problem (incorrect
default for flushroutes) addressed in a different way.
--apb (Alan Barrett)
e args is an empty string (which I hope never
happens).
--apb (Alan Barrett)
On Wed, 30 Sep 2009, Alan Barrett wrote:
> Modified Files:
> src/etc/rc.d: wpa_supplicant
> src/share/man/man5: rc.conf.5
>
> Log Message:
> Document the fact that wpa_supplicant won't start properly unless
> /usr is mounted by mountcritlocal.
I forgot to m
stem. Filesystems are then configured accordingly to their properties
> loaded from cache file.
Is /etc/zpool.cache a human-edited configuration file (in which
case, why the ".cache" name?), or is it a machine-readable database (in
which case, why is it in /etc?), or is it something else?
--apb (Alan Barrett)
these symptoms.
Does /etc/rc run at all? You could add "set -x" in strategic places
to try to figure out what's going wrong.
--apb (Alan Barrett)
buf, 1, file)
> detects EOF on an empty file.
> Fixes most of PR/41992
The changes to vfprintf.c do not seem to match the log message.
--apb (Alan Barrett)
.conf(5).)
Where was this discussed? In particular, any syntax changes in
inetd.conf should be discussed before implementation.
--apb (Alan Barrett)
On Sat, 05 Dec 2009, Adam Hamsik wrote:
> Modified Files:
> src/external/gpl2/lvm2/dist/lib/commands: toolcontext.c
>
> Log Message:
> Try to fix amd64 build.
Please describe the actual change, not just the reason for it.
--apb (Alan Barrett)
:Q}" \
The extra quotes here look wrong. ${VAR:Q} should do the correct
amount of quoting. If you need two levels of quoting for some
reason that I don't see, then use ${VAR:Q:Q}, not "${VAR:Q}".
--apb (Alan Barrett)
On Tue, 22 Dec 2009, Christoph Egger wrote:
> Log Message:
> msg.txt
Please fix the log message.
--apb (Alan Barrett)
earching recent email for
KAUTH_GENERIC_UNUSED1 returns no results for me.
> enum {
> - KAUTH_GENERIC_CANSEE=1,
> - KAUTH_GENERIC_ISSUSER
> + KAUTH_GENERIC_UNUSED1=1,
> + KAUTH_GENERIC_ISSUSER,
This change renumbers KAUTH_GENERIC_ISSUSER from 2 to 1. Was that
intentional?
--apb (Alan Barrett)
USER from 2 to 1. Was that
> > intentional?
>
> My C is rusty, if KAUTH_GENERIC_ISSUSER is 1, what is KAUTH_GENERIC_UNUSED1?
Sorry, I misread the diff. I thought that the "=1" had been removed,
which would have made KAUTH_GENERIC_UNUSED1 zero.
--apb (Alan Barrett)
t;from_mount" in the fs_spec column, and "0" in the fs_passno
column?
fs_spec="from_mount" seems to be documented only in the mount(8)
man page, not in fstab(5).
--apb (Alan Barrett)
On Fri, 12 Mar 2010, Martin Husemann wrote:
> Modified Files:
> src/distrib/sparc64/instfs: dot.profile
>
> +case $(sysctl kern.consdev 2>/dev/null) in
> + kern.consdev\ =\ ttyE*|kern.consdev\ =\ 0x5a0*)
You can simplify this using sysctl's "-n" option.
--apb (Alan Barrett)
On Mon, 19 Apr 2010, Joerg Sonnenberger wrote:
> On Sun, Apr 18, 2010 at 10:51:35AM +0000, Alan Barrett wrote:
> > Log Message:
> > Bitwise operations on signed types are well-defined if the values
> > happen to be positive, and indeed the values here were guaranteed
> &g
ake free use of $(...) elsewhere
in the NetBSD build process, so I don't see why libedit/makelist should
be an exception.
--apb (Alan Barrett)
, and that doesn't go down well on a re-build. So use
> ${INSTALL} ${COPY} instead to fix this problem.
I think that this should use ${HOST_INSTALL_FILE} -m ${SOME_MODE}.
Similarly, I think the ${CP}+${CHMOD} pairs should use HOST_INSTALL_FILE,
and the ${MKDIR} commands should use HOST_INSTALL_DIR.
--apb (Alan Barrett)
If USE_TOOLS != yes it should just use cat (or explicitly /bin/cat).
If USETOOLS != yes, then bsd.own.mk sets TOOL_CAT = cat, so it
should be fine for the makefile target to run ${TOOL_CAT}. Perhaps
postinstall should explicitly pass USETOOLS=no when it invokes
"make emit_dist_file".
--apb (Alan Barrett)
inders like
"You added a new file, but you haven't updated the sets, see
for instructions", "You added a new sysctl node but you haven't
updated sysctl(7)", etc.
--apb (Alan Barrett)
to me that funopen() creates a stream that is not associated
with a file descriptor. I assume that's what POSIX meant by "... not
associated with a file". So, I think that EBADF would be a reasonable
response.
Returning -1 without setting errno seems very wrong. Seting errno
to som
s, darkness, and most importantly:
> someone else.
What would I have to do from userland to tickle this bug?
--apb (Alan Barrett)
, then make
cleanlibdir" differ from "make cleanlibdir" ?
I suppose you wouldn't have made this change if it did not make a
difference, but what I'd like to know is, why does it make a difference?
--apb (Alan Barrett)
On Tue, 25 May 2010, Antti Kantee wrote:
> > What would I have to do from userland to tickle this bug?
>
> You need to unlink and create a file after the first namei in sys_rename
> but before VOP_RENAME runs, i.e. trigger a race condition.
Thanks.
--apb (Alan Barrett)
On Thu, 27 May 2010, David Holland wrote:
> On Thu, May 27, 2010 at 09:36:42AM +0200, Alan Barrett wrote:
> > On Thu, 27 May 2010, David A. Holland wrote:
> > > -cleandir distclean: cleanlibdir
> > > +cleandir distclean: .WAIT cleanlibdir
> >
> > I wou
On Thu, 27 May 2010, David Holland wrote:
> On Thu, May 27, 2010 at 12:26:21PM +0200, Alan Barrett wrote:
> > I think it's a bug that .WAIT affects things declared in different
> > lines, but OK, thanks for the explanation.
>
> It's a separator, not a
both the inner and outer
"echo OK" should be performed; but the test expects only the inner "echo
OK" to be performed. Similarly for the other changed tests.
--apb (Alan Barrett)
machine
it would presumably be a 36-bit type; whereas [u]int32_t would not exist
on a 36-bit machine (because C99 requires the [u]intN_t types to have
exact sizes with no padding bits).
--apb (Alan Barrett)
On Sat, 26 Jun 2010, David Holland wrote:
> I suppose the best available comprehensive solution is to use PRIu***
> garble in the code and then if necessary have autoconf figure out what
> the garble should expand to based on SIZE_MAX.
Yes.
--apb (Alan Barrett)
values in the range
from 99.95 to 99.... will be printed with one decimal place, as
"100.0".
--apb (Alan Barrett)
brain damage, but the
implementation guarantees to treat it as read-only.
--apb (Alan Barrett)
er be writable by your
program; but the implementation can still guarantee that none of the
putenv/setenv/getenv functions will write to it.
--apb (Alan Barrett)
disagrees with the comment above it. Please fix the comment.
> /* Make sure we at least double the size of the arrays. */
> - new_len = (environ_malloced_len >= 16) ?
> - (environ_malloced_len << 1) : 16;
> + new_len = environ_malloced_len >= 16 ? environ_malloced_len : 16;
--apb (Alan Barrett)
On Fri, 05 Nov 2010, Juergen Hannken-Illjes wrote:
> Log Message:
> File system snapshots are no longer experimental.
Oh really? Then why does my system crash every time I attempt to
use "dump -X"?
--apb (Alan Barrett)
PR because the feature was labeled as unstable.
I'll file one next time.
--apb (Alan Barrett)
{user} -c "/bin/sh -c 'your command here'"
works if you get the quoting right, but I still think you
need a better solution.
--apb (Alan Barrett)
On Wed, 10 Nov 2010, Christoph Badura wrote:
> On Tue, Nov 09, 2010 at 08:34:56PM +0000, Alan Barrett wrote:
> > Modified Files:
> > src/usr.bin/pathchk: pathchk.1
> >
> > Log Message:
> > Change the ironically unafe "find . -print | xargs pathchk -p&quo
Issue 6, a.k.a. IEEE Std 1003.1, 2004 edition. It's
even in Solaris.
--apb (Alan Barrett)
On Tue, 21 Dec 2010, Christos Zoulas wrote:
> Modified Files:
> src/bin/cp: cp.1 cp.c utils.c
>
> Log Message:
> Add -a archive flag. from Aleksey Cheusov
Please add a note under STANDARDS in the man page saying that
"-a" is non-standard.
--apb (Alan Barrett)
[10];
> va_list ap;
> va_start(ap, fmt);
> - (void)vsnprintf(str, len, fmt, ap);
> + (void)vsnprintf(b, len, fmt, ap);
> + (void)printf("%s\n", b);
> va_end(ap);
> }
I don't know where "len" comes from, but bad things will happen if
len > sizeof(b).
--apb (Alan Barrett)
s
explaining that the purpose of the code is to test how buffer overflows
are handled.
--apb (Alan Barrett)
> >Change group owner to operator to enable LVM locking for him.
>
> Why is /var/run not the right place for your needs?
Also, where was this discussed? If it was discussed, please
update hier(7) according to the outcome of the discussion.
--apb (Alan Barrett)
need to persist across reboot, then somewhere under
/var/run would probably be appropriate. If they are specific to lvm,
then /var/run/lvm would probably be appropriate. Anything else (such as
/var/lock or /var/run/lock) probably needs more discussion.
--apb (Alan Barrett)
that
must not be broken on reboot, /var/db may be appropriate, but then
you have to worry about the machine coming up with a different disk
configuration.
--apb (Alan Barrett)
en a long time since I
used UUCP. Given that it exists, /var/spool/lock/lvm seems fine.
> Also /var/spool/lock is not mentioned in hier.
That should be fixed.
--apb (Alan Barrett)
way to ask jot to output values from $1 to $2
in steps of 1 is "jot - $1 $2 1". I have added an example to the jot(1)
man page.
--apb (Alan Barrett)
PRIu32, some PRIu8, some
> > SCNu32, and a few cases where %u and %d were reversed.
>
> Probably you also have to add them (at least PRIu8 and SCNu32)
> into tools/compat/compat_defs.h for src/tools/disklabel.
OK, done.
--apb (Alan Barrett)
te from your preferred outcome, then you may appeal to core to
make a decision.
--apb (Alan Barrett)
RGETENV+= MKOBJDIRS=yes MKSHARE=no
> +MAKEDIRTARGETENV+= BSD_MK_COMPAT_FILE=${BSD_MK_COMPAT_FILE}
I have no opinion on whether this is the right solution to whatever the
problem is, but I can say that you don't need the ".if" guard around the
"unset MAKEOBJDIRPREFIX"; unsetting a variable that was not defined is
harmless.
--apb (Alan Barrett)
may be used by end users or third party vendors
who add custom device drivers to systems bsed on NetBSD."
--apb (Alan Barrett)
#if INTMAX_MAX < INT64_MAX
which is the reverse of what the commit log says, and it's
guaranteed by the C99 standard to never be true. (intmax_t is
guaranteed to be at least 64 bits, and is guarantted to be the
largest integral type.)
--apb (Alan Barrett)
);
+ va_end(va);
+
+ return rv;
+}
This code looks identical to that in wprintw(). Why does it need to
be duplicated under another name?
--apb (Alan Barrett)
Ah, of course.
--apb (Alan Barrett)
--apb (Alan Barrett)
g skipped
We could probably use MKUPDATE to control that. MKUPDATE is set
when you run build.sh with the "-u" option, so we know it's not a
clean build, so we do want to run "postinstall fix obsolete".
--apb (Alan Barrett)
ocal file so i could commit the changes.
could you please convert these to the add 2 lines method? thanks.
I tried to follow existing practice. There are several uses of
"NetBSD_DISABLED_filename" in other parts of the tree.
Could you refer me to an example of the add 2 lines method?
--apb (Alan Barrett)
On Mon, 25 Jul 2011, Christos Zoulas wrote:
Modified Files:
src/sbin/disklabel: disklabel.8
Log Message:
mention that we don't handle more than 2TB disks/partitions.
I think the limit's actually (2**32 - 1) sectors. With
512-byte sectors, that's one sector less than 2T
code; I don't know whether it works
in #define lines.
--apb (Alan Barrett)
twice then you get two of the many IP
addresses.
--apb (Alan Barrett)
rning.
--apb (Alan Barrett)
MONOLITHIC remains more or less equivalent to
"GENERIC with most modules preloaded".
Whether or not "available as a module" is a good reason for
removing something from GENERIC is a separate topic which I will
not consider in this message.
--apb (Alan Barrett)
fail with:
nbmtree: .: missing directory in specification
nbmtree: failed at line 1 of the specification
What do you do to provoke the problem? There's probably a fix that
doesn't involve sorting here.
--apb (Alan Barrett)
t it doesn't help.
--apb (Alan Barrett)
On Mon, 05 Sep 2011, Joerg Sonnenberger wrote:
On Mon, Sep 05, 2011 at 09:57:02AM +, Alan Barrett wrote:
Modified Files:
src/etc/mtree: Makefile
Log Message:
Use ${.OBJDIR}/NetBSD.dist.tmp instead of just NetBSD.dist.tmp.
Do we really want to add special cases like this? There
problem,
and may be willing to revert if that works out.
--apb (Alan Barrett)
On Tue, 06 Sep 2011, Alan Barrett wrote:
On Tue, 06 Sep 2011, Christos Zoulas wrote:
We definitely don't want to add such magic. Please revert. Otherwise
we should go and do this in 100's of Makefiles.
I have an idea for letting "make cleandir" deal with this problem,
a
On Sun, 11 Sep 2011, Marc Balmer wrote:
Log Message:
Install new shiny bsd.clean.mk
Thanks. I updated the set lists.
--apb (Alan Barrett)
of the libraries?
--apb (Alan Barrett)
ious error and indicates a real bug. i think having
it hidden away in just the post-boot dmesg is a poor idea.
I was confused about how log(9) worked. I have reverted the change.
Perhaps we could allow TOCONS to be passed as a flag to log(9).
--apb (Alan Barrett)
On Fri, 07 Oct 2011, Marc Balmer wrote:
Modified Files:
src/lib: Makefile
+# Lua bindings come last, they might depend on anything
+SUBDIR+= lua
Just adding it last doesn't ensure it gets built last.
You probably need a .WAIT.
--apb (Alan Barrett)
ends on the something after the last .WAIT ?
Yes, that's true.
--apb (Alan Barrett)
On Mon, 10 Oct 2011, Alan Barrett wrote:
Module Name:src
Committed By: apb
Date: Mon Oct 10 19:16:59 UTC 2011
Update of /cvsroot/src/share/zoneinfo
In directory ivanova.netbsd.org:/tmp/cvs-serv20093
Log Message:
Import tzdata2011l, from ftp://munnari.oz.au/pub
nfs layer to isolate changes made at build time.
I have not used unionfs in the past few years, but it would be a
pity to lose this functionality.
--apb (Alan Barrett)
one. I didn't even know about it.
--apb (Alan Barrett)
On Sun, 04 Dec 2011, Jean-Yves Migeon wrote:
Log Message:
Implement the register/deregister/evaluation API for secmodel(9). It
allows registration of callbacks that can be used later for
cross-secmodel "safe" communication.
Where and when was this discussed?
--apb (Alan Barrett)
1 - 100 of 237 matches
Mail list logo
