he first corrupted util.c diff and
duplicated mail.mboxfile.c diff.
OK millert@
- todd
FLEX_EXIT() is not called so flexscan() won't immediately
> terminate. This allows other syntax errors to be reported.
> The following patch applies what they have.
Let's go with the upstream patch then. OK millert@
- todd
ent the non-default settings in the comments, not
the default ones. I'm not sure we need to change this especially
since it increases the length of the comment past 80 columns.
I don't feel too strongly about this though.
- todd
> buffer. So I based it on the shell, not the kernel.
>
> Also, I deem it extremely unlikely that any configuration ever
> has a filter command which reaches over _POSIX_ARG_MAX.
OK, that's fine. We can revisit this later if system(3) is removed
from the equation.
- todd
so I don't
think you are likely to get it exactly. The kernel should enforce
that limit, not smtpd.
- todd
e each other, I'd
> say that is enough together with find(1)'s CAVEATS such that
> `find -print0 | xargs -0' does not have to me shown in EXAMPLES again.
OK millert@
- todd
On Thu, 22 Aug 2019 11:10:43 -0600, "Theo de Raadt" wrote:
> Todd C. Miller wrote:
>
> > On Thu, 22 Aug 2019 11:06:12 -0600, "Theo de Raadt" wrote:
> >
> > > Todd C. Miller wrote:
> > >
> > > > The point of that example is
On Thu, 22 Aug 2019 11:06:12 -0600, "Theo de Raadt" wrote:
> Todd C. Miller wrote:
>
> > The point of that example is to show how to safely use xargs. Since
> > find now has its own built-in xargs support perhaps we should adapt
> > the example to use that in
$ find . \( -name \*.jpg -o -name \*.gif \) -delete
- todd
don't actually
reallocate. Shouldn't the other calls that do reallocate the buffer
be converted to recallocarray(3)?
- todd
1 compat on va
> x,
> but we don't support it anymore, except on lastcomm(1) even though the kernel
> support is not present.
How about we expand the flag descriptions as follows?
- todd
Index: share/man/man5/acct.5
===
On Mon, 29 Jul 2019 13:55:32 +0100, Ricardo Mestre wrote:
> Now that this is already being used in kernel and userland the manpage for
> acct(5) needs to be updated.
Do we need to mention APLEDGE and AUNVEIL in the last sentence of
the DESCRIPTION too?
- todd
) will also benefit from reading chdir(2), in the
> same context? I don't quite see yet why that might help.
>
> I'd rather move into the opposite direction, see below.
OK millert@
- todd
We should only cross-reference the obsolete sigvec(3) function from
the signal compat manuals and sigaction(2).
This also syncs the SEE ALSO section in ualarm(3) match that of
alarm(3).
We could reference signal(3) in csh instead of sigaction(2) if
that's what people prefer.
- todd
Index
h, still too early I guess.
> What we really want to check is:
>
> if (itp->it_value.tv_sec > 0 ||
> (itp->it_value.tv_sec == 0 && itp->it_value.tv_usec > 0))
>
> which is nearly equivalent to the expansion above. The expansion
> above ignores the sign of itp->it_value.tv_usec, so, not exactly the
> same.
Right.
- todd
o the timerisset() call appears to be superfluous. I see no case
where "itp->it_value.tv_sec >= 0" is true that timerisset(&itp->it_value)
is not also true.
Am I missing something?
- todd
for chipping away at relative timeouts. Then we could write
>
> if (!timerisexpired(&itp->it_value))
>
> ... which is precisely what we want to express.
That would make it easier to read for sure.
- todd
ing easier to read:
if (itp->it_value.tv_sec >= 0 && itp->it_value.tv_usec > 0)
(yes, I know the tv_usec could just be != 0).
OK millert@ either way,
- todd
On Thu, 25 Jul 2019 12:00:48 +0200, Alexander Bluhm wrote:
> Do we want unveil violators in the daily mail? We can turn it off
> if we get too many false positives.
I think so. If it becomes annoying we can turn it off by default.
- todd
e. OK millert@
- todd
just fire less often. We could add a warning about it
> in CAVEATS in ping.8...
I'm fine with removing that code; I don't think a CAVEAT in the
manual is needed.
- todd
sh output at the end of writing a page in raw mode (-B or -b
> flags).
OK millert@ if you fix the whitespace (should use tabs for indent
and match the indent of the preceding lines).
- todd
place the ether_isbcast, ether_isequal and
ether_cmp macros in net/if_gre.c as well.
- todd
need for endpwent() there.
- todd
fy one. Callers need to make their own struct with those
elements. Any code that relies on "struct mymsg" in sys/msg.h will
not be portable.
- todd
d
our own man page).
- todd
ps://pubs.opengroup.org/onlinepubs/9699919799/
That all makes sense to me. Checking for ngrp<0 in setgroups()
makes things easier to understand compared to the implicit cast to
uint and relying on the result to be >NGROUPS_MAX.
- todd
On Tue, 02 Jul 2019 09:43:13 -0600, "Theo de Raadt" wrote:
> Yes, that is a problem. Same problem applies to two other files;
> here are the diffs I propose.
Looks good. OK millert@
- todd
On Tue, 25 Jun 2019 23:57:24 +0200, Sebastian Benoit wrote:
> Well, if you do that, i think some callers need a look, for example
That change was specific to F_SETLK so the callers you reference
are not affected. However, I did miss F_GETLK, which needs the
same treatment.
- todd
Index:
On Tue, 25 Jun 2019 09:18:28 -0600, Todd C. Miller wrote:
> The ERRORS section could be clearer as to what error you get when
> you try to lock a socket/fifo.
There is no need to change fcntl.2 or lockf.2, they already document
the behavior under EINVAL. The only outlier now is flock.2
On Tue, 25 Jun 2019 09:18:28 -0600, "Todd C. Miller" wrote:
> The ERRORS section could be clearer as to what error you get when
> you try to lock a socket/fifo.
Actually, for fcntl(2), we should be returning EINVAL, not EBADF
when trying to lock a non-file. Let's fix that f
The ERRORS section could be clearer as to what error you get when
you try to lock a socket/fifo.
- todd
Index: lib/libc/gen/lockf.3
===
RCS file: /cvs/src/lib/libc/gen/lockf.3,v
retrieving revision 1.13
diff -u -p -u -r1.13 lockf.3
ad the diff. The placement in that diff is fine.
- todd
ould just place that bit either before the options are described
or after the options in an "Execution environment" sub-section.
- todd
> 3. With keepenv, other variables are retained, but the above are still reset.
>
> 4. Possible to override this with setenv.
>
> This is much more consistent and predictable.
That does seem better. The user can always preserve some or all
of those variables via setenv. OK millert@
- todd
recovery since you'd
need to compute the required length. One of the reasons for
strlcpy/strlcat was to avoid needing to compute lengths since that
is often a source of bugs.
In any event, the strlcpy/strlcat interface has been around for
over 20 years now. You can't change an API like that without
breaking all the code that uses it.
- todd
ack"
though you should escape the '.' if you want it to match literally.
- todd
e patch sent
> out here because it would make the patch unreadable.
OK millert@
- todd
On Sat, 11 May 2019 00:30:35 +0200, Ingo Schwarze wrote:
> While here, i noticed ugly preprocessor macros.
> Let's make our future life easier by unifdefing a bit.
> When compiling with -g0, there is no object change.
No objection, OK millert@
- todd
On Fri, 10 May 2019 23:06:21 +0200, Ingo Schwarze wrote:
> The following patch preserves the parsing behaviour
> and correctly stores the number of seconds into tm_gmtoff.
That one looks correct. OK millert@
- todd
ieve this handles the "[+-]hh" form.
- todd
ve, improve and add
> comments. Also, the comment above pshift() is simply wrong.
> The function does not discard a specific number of characters.
Looks good. OK millert@
- todd
On Thu, 02 May 2019 21:36:52 -0700, Andrew Hewus Fresh wrote:
> In any case, I'll probably commit this slightly cleaned up version
> tomorrow as it's to late right now to make sure I didn't clean it too
> far.
OK millert@
- todd
ugh I haven't
tested it yet.
- todd
Index: gnu/usr.bin/perl/cpan/Pod-Perldoc/lib/Pod/Perldoc/ToMan.pm
===
RCS file: /cvs/src/gnu/usr.bin/perl/cpan/Pod-Perldoc/lib/Pod/Perldoc/ToMan.pm,v
retrieving revision 1.8
diff -u -p -u -r1.8 T
ontains the following:
implementations may restrict the permissions, either by clearing
the file mode bits or setting them to the value S_IRUSR | S_IWUSR.
So your diff is OK millert@
- todd
That diff accidentally removed the unlink() but I'm not convinced
it is a good idead either way as it introduces a race been open and
unlink where another user could open the file, depending on the
caller's umask.
- todd
On Thu, 25 Apr 2019 09:34:47 -0600, Todd C. Miller wrote:
> I think a better approach is to add an internal version of mkstemp(3)
> that takes a permission flag. We can then pass in DEFFILEMODE for
> the permission bits like stdio and avoid the umask and fchmod
> entirely.
Somethi
I think a better approach is to add an internal version of mkstemp(3)
that takes a permission flag. We can then pass in DEFFILEMODE for
the permission bits like stdio and avoid the umask and fchmod
entirely.
- todd
On Tue, 23 Apr 2019 14:52:09 +0200, Ingo Schwarze wrote:
> That wouldn't be me. Either way, an additional OK to import would be
> appreciated.
I like this too. OK millert@ for importing it.
- todd
;t clear the screen by default on ^L. Other shells
like bash, zsh, and tcsh do. I don't object to making it the default
but as I'm not a ksh user I'll defer to those who are.
- todd
drawbacks and the missing features will be
> readily implemented in time for the 6.6 release.
Will the bpf JIT changes be done in time for 6.6? I have no doubt
that "pfctl -p /dev/bfp" can be made to work in time but for a truly
performant firewall we will need bpf JIT.
- todd
On Mon, 25 Mar 2019 21:44:45 +0100, Anton Lindqvist wrote:
> My guess is that flag actually refers to atflags.
Yes, looks like copy pasta from fchmodat. OK millert@
- todd
On Fri, 22 Mar 2019 11:16:00 -0500, Scott Cheloha wrote:
> ... it says the modification time is a time_t. Which means we only
> have seconds, not subseconds.
Fair enough. The protocol predates common availability of nanosecond
file timestamps.
- todd
On Fri, 22 Mar 2019 11:04:05 -0500, Scott Cheloha wrote:
> I always forget about the special value shortcut for utimes(2) et al.
>
> This is equivalent/simpler/more portable.
That looks good but I wonder why we are not preserving the nanosecond
mtime by using st_mtim?
- todd
n the distribution (though it is in the nmh repo).
- todd
mail" for argv[0] in its spost utility. It probably got that
from mail(1) originally.
- todd
On Tue, 19 Mar 2019 10:33:07 +0200, Lauri Tirkkonen wrote:
> ping - doesn't look removed yet :)
Committed.
- todd
Wouldn't it be less error-prone to make escape_char u_char instead
of int?
- todd
keyboard.
> ~tilde is certainly annoying when it's three key presses alone,
> and then you mostly get only one shot at trying..
The cu from Taylor UUCP uses -E for this (it uses -e to specify
even parity). If we are going to support this, I think it makes
sense to be as compatible as possible with other systems.
- todd
ondition
> because this "else if" clause is supposed to handle non-printable
> single-byte characters.
>
> 4. The call to the bad function is_ubin_char() intends to handle
> non-printable Unicode characters, so the right function to use
> is simply !iswprint(3) from .
This looks like an improvement to me. OK millert@
- todd
end-mail" as far as I know.
At this point I consider it a historical oddity that is safe to
remove.
- todd
On Tue, 26 Feb 2019 02:39:14 +0100, Ingo Schwarze wrote:
> In that case, let's just use an index rather than a pointer;
> diff otherwise unchanged.
OK millert@
- todd
On Mon, 25 Feb 2019 19:43:36 +0100, Ingo Schwarze wrote:
> Todd C. Miller wrote on Mon, Feb 25, 2019 at 09:45:12AM -0700:
>
> > On Mon, 25 Feb 2019 12:39:41 +0100, Ingo Schwarze wrote:
>
> >> Index: line.c
> [...]
> >> @@ -469,11 +469,10 @@ in_ansi_esc_seq(v
On Mon, 25 Feb 2019 12:39:41 +0100, Ingo Schwarze wrote:
One question inline.
- todd
> Index: line.c
> ===
> RCS file: /cvs/src/usr.bin/less/line.c,v
> retrieving revision 1.23
> diff -u -p -r1.23 line.c
> --
CSI).
That makes the most sense to me. The simpler we can make less(1)
the better.
- todd
A few weeks ago I noticed that firefox tabs were getting killed for
running afoul of pledge(2). It seems that the problem was some calls to
shmget(2) from the X swrast_dri.so lib that seem to have come from the
new mesa code that was recently imported. Since the shm syscalls aren't
covered by any p
; lf_split().
OK millert@
- todd
rence with my email dirs ;-)
- todd
On Tue, 19 Feb 2019 23:48:03 +0100, Klemens Nanni wrote:
> Given this feedback and the fact that the majority of implementations
> already behave that way: any OKs or objections for my diff?
OK millert@
- todd
The diff below teaches the lldb assembly inspector to skip over the
retguard instrumentation when traversing function prologues.
ok?
diff --git
a/gnu/llvm/tools/lldb/source/Plugins/UnwindAssembly/x86/x86AssemblyInspectionEngine.cpp
b/gnu/llvm/tools/lldb/source/Plugins/UnwindAssembly/x86/x86As
ms to apply to the OpenBSD copy of the cod
> e?
>
> https://github.com/freebsd/freebsd/commit/95b1feb22b6da0549367ed0023308f0abfa
> 8086b#diff-d7c26714f9432399b202eefcedb97491
Here's a diff relative to our regcomp.c.
- to
On Mon, Feb 04, 2019 at 04:53:36PM +0100, Sebastien Marie wrote:
> Hi,
>
> Recently, devel/llvm (the port) has copied the -msave-args diff from
> base, and it resulted lang/rust to segfault.
>
> Since, the change has been backouted, but I continued searching the root
> cause as the diff is on base
On Tue, Jan 22, 2019 at 02:47:10PM +0900, YASUOKA Masahiko wrote:
> Hi,
>
> Currently gdb "target kvm bsd.0.core" on amd64 doesn't work. It can't
> read the registers and can't follow the stack frames. The diff
> follows fixes those problems.
>
> ok? comment?
Sorry for the delay getting to this.
This was something I meant to fix years ago after a libc major
version bump but forgot about.
The use of int for gl_pathc and gl_offs is historic. This is a BSD
interface and POSIX changed them to size_t when it was standardized.
- todd
a long time since
> the typical font rendered them symmetrically.
OK millert@
- todd
for job
control support.
- todd
ATS. Less bloat everywhere, better overview in the one critical
> place.
>
> Does that make sense?
Sure, that would be similar to how we list signal safe functions
in sigaction(2).
- todd
here people would see it, but I'm not sure
there really is one.
- todd
Don't we have the same issue with printf(3) and scanf(3)?
For example, in src/lib/libc/stdio/vfprintf.c:1022
decimal_point = nl_langinfo(RADIXCHAR);
Also, anything using strtod(3), including scanf(3), is likewise
affected by gdtoa's use of localeconv()->decimal_point.
- todd
;ve had that happen in the (now rather distant) past.
- todd
The main difference with -S is that you end up with two copies of
the target file on disk for a short period of time. That was a
bigger deal back when -S was added. Today, it probably doesn't
matter.
- todd
On Thu, 03 Jan 2019 12:51:48 +0100, Martijn van Duren wrote:
> On 1/3/19 11:08 AM, leo_...@volny.cz wrote:
> > Alright, these ones get rid of is_stdin altogether. Isn't it better this
> > way?
> >
> Looks OK to me, anyone else want to OK this?
OK millert@
- todd
d special
> handling of empty values, leaving a diff like the one below.
OK millert@
- todd
h me. I definitely think it should be installed
somewhere. I found it odd that I had to look for mandoc.css in the
source tree when I needed it.
- todd
- todd
This has the side effect of closing stdin on EOF but I don't think
that is a problem. OK millert@
- todd
On Wed, 05 Dec 2018 22:52:13 -0500, "Ted Unangst" wrote:
> Seen in the wild. Alias for ! that's friendlier to the shell.
OK millert@
- todd
Are you sure these checks cannot be hit? For example:
% ifconfig lo0 group jkhaksudhkashdkashdkahdksahdkjhakjshdaksjhd
ifconfig: setifgroup: group name too long
sure looks like the check gets triggered.
- todd
On Sat, 10 Nov 2018 18:43:11 +0100, Anton Lindqvist wrote:
> Make use of the new LFPRINT() debug macro.
OK millert@
- todd
_cmds[${#_cmds[*]}]="${_cmd#!}"
| > + ;;
| > + bridge)
| > + # Skip bridge in the installer.
| > return
| > ;;
| > *) _cmds[${#_cmds[*]}]="ifconfig $_if ${_c[@]}"
| >
| >
| > Cheers
| >
--
Todd T. Fries . http://todd.fries.net/pgp.txt . @unix2mars . github:toddfries
On Mon, 05 Nov 2018 08:12:57 +0100, Anton Lindqvist wrote:
> Start tracing struct flock. I've been using this diff during lockf
> development.
OK millert@
- todd
On Thu, Nov 01, 2018 at 12:58:36PM +0100, Matthieu Herrb wrote:
> Hi,
>
> The XDM-AUTHORIZATION-1 protocol is currenly disabled in xenocara.
> It makes sense: this auth protocol doesn't work with IPv6, is relying
> on weak DES encryption and only useful with XDMP which isn't supported
> by xenodm.
On Fri, 02 Nov 2018 00:45:06 +0100, Klemens Nanni wrote:
> On Thu, Nov 01, 2018 at 06:37:15PM -0400, Ted Unangst wrote:
> > yes, this is how it should be. please fix group lookup as well.
> Was in the queue for a separate diff, here it goes together.
Looks good. OK millert@
- todd
On Thu, 01 Nov 2018 21:04:03 +0100, Klemens Nanni wrote:
> Here's a diff that tries UIDs only if username lookup fails.
OK millert@
- todd
On Wed, 31 Oct 2018 23:38:58 +0100, Klemens Nanni wrote:
> New diff free of side-effects in filteruser(), rest unchanged.
OK millert@
- todd
effect in filteruser(), see below.
Otherwise looks good.
- todd
> Index: top.c
> ===
> RCS file: /cvs/src/usr.bin/top/top.c,v
> retrieving revision 1.94
> diff -u -p -r1.94 top.c
> --- top.c 5 Oct 2018 18:56:57 -
OK millert@
- todd
; Since we already call pledge(2) before, twice, we need to add "unveil" promis
> e
> to both of them, and finally call pledge(2) once again with the needed promis
> es
> except "unveil".
OK millert@
- todd
With your patch the mount point is offset by a space from the
"Mounted on" header.
- todd
On Thu, 25 Oct 2018 11:19:34 +0100, Ricardo Mestre wrote:
> The code path were we pass `pathname' in the arguments is already limited
> with pledge(2), but since we know exactly what it is then we can go further
> and also unveil(2) it with read permissions.
OK millert@
- todd
On Wed, 24 Oct 2018 08:05:11 +0100, Ricardo Mestre wrote:
> The only file that spamlogd needs to access after calling pledge is
> PATH_SPAMD_DB, so unveil it with O_RDWR permissions.
Looks good. OK millert@
- todd
601 - 700 of 1653 matches
Mail list logo
