On Mon, 29 Jul 2019 18:14:14 +0100, Ricardo Mestre wrote:
> Sorry, I'm not too good with words for manpages.
>
> Is this good enough? I'll commit the AUNVEIL bit separately anyway in a few
> minutes.
>
> And while we are here what about ACOMPAT? It was used for PDP-11 compat on va
> x,
> but we do
On Mon, 05 Aug 2019 17:31:51 +0900, YASUOKA Masahiko wrote:
> The diff basically is to do the same thing which is done on the
> upstream, but it also replaces h_malloc() and tok_malloc() which the
> upstream didn't replace yet.
This only changes the reallocarray(3) calls that don't actually
reall
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 instead.
We can also list the -delete method as an alternative. E.g.
$ find . \( -name \*.jpg -o -name \*.gif \) -exec rm {} +
or
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
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 19:20:53 +0200, Klemens Nanni wrote:
> I concur. We can at least showcase the "batching" version that passes
> multiple arguments to rm(1) while also mentioning `-delete'.
>
> Both find(1)'s `-print0' and xarg(1)'s `-0' reference each other, I'd
> say that is enough together w
On Mon, 02 Sep 2019 21:15:23 +0200, Martijn van Duren wrote:
> https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/limits.h.html
> {_POSIX_ARG_MAX}
> Maximum length of argument to the exec functions including environment da
> ta.
> Value: 4 096
Note that this is the minimum value PO
On Tue, 03 Sep 2019 17:11:22 +0200, Martijn van Duren wrote:
> I choose this value because I hit the maximum command length of the
> shell before. This way I'm somewhat confident that the shell doesn't
> do something weird with my command if we ever overflow it's internal
> buffer. So I based it o
On Tue, 10 Sep 2019 12:59:51 +0200, Solene Rapenne wrote:
> I looked at /etc/examples/sysctl.conf on an amd64 system and found 2
> things:
>
> - file refers to sysctl(3) and sysctl(8). sysctl(3) doesn't exists but
> sysctl(2) exists, I think we want a 2
Yes, sysctl(3) was renamed to sysctl(2) s
On Sat, 21 Sep 2019 20:21:12 +0800, Michael Mikonos wrote:
> When I remembered to check if upstream flex had fixed the sf_pop()
> issue I found their fix was a little different:
> https://github.com/westes/flex/commit/9ba6e5283efd2fe454d3bc92eca960b3ebd9129
> 4
>
> Notably, FLEX_EXIT() is not call
On Wed, 25 Sep 2019 15:23:06 +0200, Martijn van Duren wrote:
> I talked to gilles@ about this and he's somewhat inclined to agree with
> me that this is a TEMPFAIL situation. Even if the directory where the
> mail is supposed to be delivered is permanently not accessible I'd
> argue that we should
On Mon, 07 Oct 2019 15:25:13 +0100, Stuart Henderson wrote:
> Attempting to build Firefox on aarch64 needs at least 4.5GB but I think
> it would be better to just sync with amd64. OK?
Makes sense. OK millert@
- todd
On Wed, 16 Oct 2019 10:23:52 -0500, Scott Cheloha wrote:
> Tracking the process starting time as an uptime fixes the classic
> "init(8) started in 1969" bug in ps(1) when your CMOS battery dies.
>
> In general it lets us track how long a process has been running
> correctly regardless of whether o
On Sat, 26 Oct 2019 12:15:33 +0100, Stuart Henderson wrote:
> The way these files are supposed to work is that you set the system
> clock to the time with leap-seconds included (UTC+leap, or TAI-10) and
> copy the entire "right" set of files to the main zoneinfo directory
> (upstream provides them
On Fri, 04 Feb 2022 20:46:48 -0800, Greg Steuck wrote:
> How do people feel about shipping the minimal UBSan runtime library[1]
> in the base system? It takes very little to build (Makefile + a few
> ifdefs that both jca@ and I hacked together). The library is tiny and
> useful enough for finding
On Sat, 05 Feb 2022 23:06:32 +1100, Jonathan Gray wrote:
> drop please from manual pages excluding third party parts of tree
> suggested by multiple documentation style guides
OK millert@
- todd
On Sat, 05 Feb 2022 14:38:14 +0100, Christian Weisgerber wrote:
> Anybody?
> The commit history of ypldap does not suggest a single responsible
> person I could pin down.
OK millert@
- todd
Since the input is opened read-only I don't see the point in checking
the fclose() return value. However, if you are going to do so, you
might as well combine it with the ferror() check. E.g.
if (ferror(fp) || fclose(fp) == EOF) {
warn("%s", name);
status
On Sun, 06 Feb 2022 23:41:26 -0600, Scott Cheloha wrote:
> > On Feb 6, 2022, at 20:07, Todd C. Miller wrote:
> >
> > Since the input is opened read-only I don't see the point in checking
> > the fclose() return value. However, if you are going to do so, you
>
On Tue, 08 Feb 2022 10:40:00 -0600, Scott Cheloha wrote:
> We don't need "rpath" if we're only processing the standard input.
Sure. OK millert@
- todd
On Tue, 08 Feb 2022 19:37:26 -0600, Scott Cheloha wrote:
> - pledge(2) initially with "stdio rpath" at the top of main().
> We know we need to read a file at this point but don't yet
> know which one.
>
> - pledge(2) down to "stdio" after we have opened the file
> in question and called fsta
On Wed, 09 Feb 2022 11:35:58 +0100, Theo Buehler wrote:
> On clang we can use __has_feature(), but that doesn't exist on
> gcc which defines __SANITIZE_ADDRESS__ if it compiles with
> -fsanitize=address.
>
> This doesn't warn on sparc64 and works in my test setups.
It's a little ugly but since th
On Thu, 10 Feb 2022 09:13:25 +0100, Theo Buehler wrote:
> This is purely cosmetic. I did some testing on fedora which ships with
> btrfs by default. btrfs is special in that df -i and other tools always
> report 0 inodes. As a consequence, each rpki-client run prints the disk
> space warning, whic
On Fri, 11 Feb 2022 08:21:38 +0100, alf wrote:
> there seem to be some superfluous words in
> lib/libc/gen/statvfs.3 and sys/sys/statvfs.h .
> "(unit f_frsize)" can be removed since in the man page
> it is explicitely mentioned directly below:
> " The fields of type fsblkcnt_t are reported in unit
On Tue, 15 Feb 2022 20:01:52 +0100, Florian Obser wrote:
I think you need that to be:
/* MUST delete trailing NUL, per RFC 2132 */
slen = dho_len;
while (slen > 0 && p[slen - 1] == '\0')
slen--;
to avoid underflow if the string happens to consist entirely
On Tue, 15 Feb 2022 20:18:45 +0100, Florian Obser wrote:
> Also fixes a whitespace issue while here.
>
> if (dho_len < 1)
> goto wrong_length;
>
> is redundant now, but I want to keep the pattern of checking the length
> right after identifying the option.
Right, makes sense.
On Sun, 13 Feb 2022 12:07:31 -0800, Greg Steuck wrote:
> I noticed that despite the OKs the code didn't submitted. Should we
> revive this diff now and continue in the tree?
>
> I have vested interest: I don't want to fix up lang/ghc test suite which
> is full of seq(1).
Here's an updated diff th
On Wed, 23 Feb 2022 00:04:07 +0100, Theo Buehler wrote:
> Hard to hit. Nevertheless it seems better to error out than to
> crash in strcmp()
Sure. OK millert@
- todd
On Sat, 26 Feb 2022 11:16:44 -0600, Matthew Martin wrote:
> Anton spotted a doas regression failure in t-run-keepenv-path after the
> change to doas for LOGIN_SETALL. Since that test runs doas in a chroot
> and the setup does not create a login.conf, login_getclass in
> login_cap.c will return a l
On Wed, 02 Mar 2022 19:26:46 +0100, Theo Buehler wrote:
> I think it's easier to use a single snprintf call. Also drop a few
> unnecessary parentheses.
Seems like an improvement. OK millert@
- todd
On Thu, 03 Mar 2022 13:13:16 +0100, Solene Rapenne wrote:
> The following diff adds support for -k flag to keep the input file for
> gzip / compress when compressing, and the input file (the compressed
> one) for gunzip / uncompress
>
> This will improve uses cases like: zcat -f "${file}" > "${fil
On Thu, 03 Mar 2022 15:11:13 +, Miod Vallat wrote:
> > I think this makes sense if only for better GNU gzip compatibility.
> > OK millert@
>
> But does the `-k' flag needs to be added to compress(1) too?
No, it just makes usage() slightly more complicated.
But that diff was missing an update
On Sun, 06 Mar 2022 02:58:30 +0100, Jeremie Courreges-Anglas wrote:
> I'm not sure what you mean here. Solene's diff added -k to both
> compress(1) and gzip(1) (and their uncompressor counterparts).
> Adding -k to gzip/gunzip only would indeed make the usage() slightly
> more complicated.
>
> So
On Wed, 09 Mar 2022 19:20:08 -0600, Scott Cheloha wrote:
> The strdup(3) implementation in libc uses memcpy(3), not strlcpy(3).
>
> There is no upside to using strlcpy(3) here if we know the length of
> str before we copy it to the destination buffer.
Sure, using memcpy() here is fine since the l
On Sun, 13 Mar 2022 13:26:13 +0100, Jeremie Courreges-Anglas wrote:
> Works fine. Here's an updated diff with suggestions:
> - "k" was not completely removed from compress's struct compressor opt
> string, and was not needed in null_method
> - try to keep the *flag variables ordered
> - rework
On Sun, 13 Mar 2022 12:02:03 -0500, Matthew Martin wrote:
> Ignoring -L which already honors rtable, su has three cases:
> -l (asme=0 asthem=1)
> -m (asme=1 asthem=0)
>(asme=0 asthem=0)
>
> -l should honor rtable; I am not sure about the other two. I think the
> least suprising would be fo
I think it makes sense to accept LOGIN_SETRTABLE for setclasscontext(3)
as well. Currently, it will be cleared out of the flags.
- todd
Index: lib/libc/gen/login_cap.3
===
RCS file: /cvs/src/lib/libc/gen/login_cap.3,v
retrieving re
On Wed, 23 Mar 2022 15:53:21 +0100, Alexander Bluhm wrote:
> Hi,
> is a big switch and when looking at a break you need more context
> to see where it jumps to.
>
> I would like to use goto fail consistently to leave the big switch.
> break is used for inner switches and loops.
This makes it easi
On Wed, 23 Mar 2022 16:59:06 +0100, Otto Moerbeek wrote:
> This is a bug that dlg reported last week. Serendepity or not? :-)
>
> This is my diff that uses an approach I like a litle bit better.
Since client_peer_init() is always called after new_peer() there's
no reason I can see for query to be
On Sun, 27 Mar 2022 15:05:00 +0200, Sebastien Marie wrote:
> The following diff removes vnode_hold_list, vnode_free_list extern references
> ,
> and `struct freelst` definition from sys/vnode.h
>
> `struct freelst` definition (TAILQ_HEAD) is moved where used, inside vfs_subr
> .c.
OK millert@
On Sun, 10 Apr 2022 13:24:15 +0200, Jeremie Courreges-Anglas wrote:
> The diff below keeps the same output by default and outputs the same as
> GNU gzip(1) is -k is given.
OK millert@
- todd
On Thu, 14 Apr 2022 17:52:37 -0500, joshua stein wrote:
> login_fbtab(3) supports wildcards but only for every file in a
> directory (/path/*).
>
> This makes it use glob(3) so it can also support more specific
> wildcards like /path/file*
Yes, it is better to use glob(3) than something custom.
On Fri, 15 Apr 2022 13:12:03 -0500, joshua stein wrote:
> Thanks, both applied.
Looks good to me but needs a man page update.
- todd
On Fri, 15 Apr 2022 13:28:33 -0500, joshua stein wrote:
> Anyone want to bikeshed this language?
I think it is more helpful to refer to glob(7) than glob(3).
Perhaps something like this.
- todd
Index: share/man/man5/fbtab.5
===
RC
On Tue, 19 Apr 2022 11:41:10 +0200, Martin Vahlensieck wrote:
> There is no need to duplicate options->send_env[i] only free it
> in all cases. Just use options->send_env[i] directly.
Good catch, match_pattern() takes const char * arguments so
there is no danger there.
- todd
On Tue, 19 Apr 2022 11:42:52 +0200, Martin Vahlensieck wrote:
> malloc(3) and friends require stdlib.h, SIZE_MAX requires stdint.h.
You are correct. The other xmss files get these includes via
xmss_commons.h, but ssh-xmss.c does not.
- todd
On Tue, 19 Apr 2022 11:43:35 +0200, Martin Vahlensieck wrote:
> Neither openssl/evp.h nor openssl/hmac.h are required.
Right.
- todd
On Tue, 19 Apr 2022 21:30:25 -0500, Matthew Martin wrote:
> On Thu, Apr 07, 2022 at 06:11:45PM -0500, Matthew Martin wrote:
> > const the termp and winp arguments for openpty and related. This matches
> > the prototypes for openpty and forkpty in glibc and musl libc.
>
> ping; has an ok from tb@ [
On Wed, 20 Apr 2022 09:18:20 -0500, joshua stein wrote:
> I like it. Here's the full diff:
OK millert@
- todd
On Thu, 21 Apr 2022 09:35:44 -0700, Jeremy Mates wrote:
> The cause is an unguarded use of the NULL output pointer. I am pretty
> sure an .exrc cannot cause this condition (map rhs requires
> something, not nothing) only recompiling with a NULL output string
> for some command.
>
> One fix is to g
Another thing I noticed while testing the fix is that in cl_term_init()
while processing the command mapping there is a missing NULL check.
However, the check _is_ present for the input mapping loop. The
following diff makes the two loops consistent.
I'm not sure whether this is worth doing since
[I originally sent this a year ago but didn't follow up.]
This is consistent with vim's expandtab behavior.
>From nvi2 (Craig Leres):
https://github.com/lichray/nvi2/commit/7d3f43559026e0927032a105b217850feaefec66
I know most of us don't use expandtab but it is helpful when dealing
with a codeba
On Fri, 22 Apr 2022 17:38:56 +0200, Theo Buehler wrote:
> I would use an extra variable. The simplification alone is a good
> argument to do it:
Yes, that is nicer. OK millert@
- todd
On Thu, 28 Apr 2022 20:54:02 -0500, Scott Cheloha wrote:
> Once again, using nanosleep(2) here to print the stats periodically is
> flawed. The period will drift. Using setitimer(2)/sigsuspend(2) is
> better.
Yes, I agree that an interval timer is a better fit.
> While here:
>
> - We don't nee
On Sun, 01 May 2022 00:48:17 -0400, "Ted Unangst" wrote:
> I don't think we need to concern ourselves with cross awk compatibility here.
>
> Despite the misleading comment, /usr/bin/awk supports toupper.
This must have been to support very old versions of awk (pre-nawk)
since toupper and tolower
On Mon, 02 May 2022 18:39:17 +0200, Martin Pieuchot wrote:
> Let's simplify the existing logic and use a single list for inactive
> pages. uvmpd_scan_inactive() already does a lot of check if it finds
> a page which is swap-backed. This will be improved in a next change.
That looks fine to me.
On Mon, 02 May 2022 20:39:07 -0500, Scott Cheloha wrote:
> Whenever I block signals, deraadt@ rises up out of the floorboards and
> says "I hate masking signals, don't do that."
>
> ... but if millert@ is still fine with the attached patch, which does
> sigprocmask(2), I'll go ahead with it.
>
> >
On Tue, May 03, 2022 at 10:37:36PM -0500, Matthew Martin wrote:
> The function is already marked __dead in passwd.c, so appears to just be
> an oversight.
Committed, thanks.
- todd
On Fri, 13 May 2022 19:33:58 -0700, Philip Guenther wrote:
> Diff below converts the last three such uses** and then updates the symbol
> naming for fseek to block attempts to reference it in the future so uses
> don't accidentally crawl back in.
OK millert@
- todd
On Sat, 14 May 2022 09:31:26 -0500, Scott Cheloha wrote:
> ok?
OK millert@
- todd
Our tar's -O flag is only used when creating an archive, it is
unused for extraction. I'd prefer that we use the same option
letter as GNU tar and bsdtar for this.
- todd
On Thu, 02 Jun 2022 14:36:16 +0200, Jan Stary wrote:
> That results in the above. What obsolete options format
> is this trying to accomodate? The manpage doesn't say -
> the options it describes are perfectly getopt()-likable.
> Looking at the CVS log, this was already "obsolete"
> in the origina
On Thu, 02 Jun 2022 07:43:15 -0600, "Theo de Raadt" wrote:
> Hmm, but consider these cases
>
> dump home
>
> or
>
> mkdir 0af
> dump 0af
>
> or
>
> cd /dev && dump rsd0a
True, those would not be handled but isn't the most common usage
to pass a fully-qualified path or a device name? The biggest
On Thu, 02 Jun 2022 07:54:02 -0600, "Theo de Raadt" wrote:
> I'm fine with a / check, but it also needs documenting. While there can't
> we say at least one option must be supplied?
How about this?
- todd
Index: sbin/dump/dump.8
Instead of the existing alarm() mechanism, how about using
timeout_connect() from usr.bin/nc/netcat.c?
- todd
On Sun, 10 Jul 2022 20:58:35 -0900, Philip Guenther wrote:
> Three thoughts:
> 1) Since stdio errors are sticky, is there any real advantage to checking
> each call instead of just checking the final fclose()?
Will that really catch all errors? From what I can tell, fclose(3)
can succeed even if
On Wed, 31 Jan 2018 20:44:10 +0100, Landry Breuil wrote:
> the default etc/syslog.conf has a commented out example for by-prog
> logging, but nothing for by-host logging. I fighted a bit with it; so
> why not providing an example in the EXAMPLES section of the manpage ?
I recently did the same so
On Thu, 01 Feb 2018 01:55:18 +0200, Kapetanakis Giannis wrote:
> On 01/02/18 00:06, Todd C. Miller wrote:
> >
> > Shouldn't this be:
> >
> > # Log everything coming from host bastion to a separate file
> > ++bastion /
OK millert@ for that version.
- todd
On Mon, 05 Feb 2018 09:21:32 +1300, Theo Buehler wrote:
> The load_user() function gets a file descriptor from process_crontab().
> It fdopen()s it directly and fclose()s the resulting stream. Then
> process_crontab() closes the stream a second time before exiting.
>
> Since crontab_fd is not load
On Mon, 12 Feb 2018 10:08:25 -0600, Scott Cheloha wrote:
> I'm unfamiliar with this code, but assigning the microsecond value
> without normalizing it for the timeval looks off.
That makes sense. There's nothing preventing the user from using
values > 100. OK millert@
- todd
OK millert@
- todd
On Sat, 17 Feb 2018 20:10:45 +0100, Mark Kettenis wrote:
> Not that vmmap is a much better name, but that's what we use on other
> non-pmap-direct architectures.
Consistency is good, OK millert@
- todd
On Thu, 22 Feb 2018 13:50:13 -0600, Scott Cheloha wrote:
> I think setjmping from a signal handler to put a time limit on
> pclose is too magical, especially when the alternative doesn't
> require much more code.
Agreed.
> But I do think putting a time limit on our wait for wall to do its
> job
On Thu, 22 Feb 2018 15:06:04 -0600, Scott Cheloha wrote:
> Could that difference effect the behavior of the program in practice?
I don't think so.
> Attached diff is using the file stream functions still, for comparison.
>
> But the dprintf diff feels more natural; keeping the stream functions
>
On Mon, 26 Feb 2018 19:12:20 -0600, Scott Cheloha wrote:
> Reading the latest POSIX description for times(3):
>
> > Upon successful completion, times() shall return the elapsed
> > real time, in clock ticks, since an arbitrary point in the past
> > (for example, system start-up time). This point d
On Wed, 28 Feb 2018 09:46:30 -0600, Scott Cheloha wrote:
> I want to avoid giving the reader even the slightest impression that
> the return value from times(3) can be used for anything but real-time
> interval measurement. Mentioning that the value is relative to the
> system start time seems to
On Wed, 28 Feb 2018 10:23:24 -0600, Scott Cheloha wrote:
> I know it's not customary to put extra details in RETURN VALUES, but
> times(3) is a bit unwieldy. How does the attached read?
OK by me.
- todd
On Thu, 15 Mar 2018 17:23:24 +0100, Patrick Wildt wrote:
> diff --git a/gnu/usr.bin/binutils-2.17/include/obstack.h b/gnu/usr.bin/binuti
> ls-2.17/include/obstack.h
> index 88c2a264adc..8839c48e95f 100644
> --- a/gnu/usr.bin/binutils-2.17/include/obstack.h
> +++ b/gnu/usr.bin/binutils-2.17/include
On Fri, 16 Mar 2018 14:23:29 -0500, Scott Cheloha wrote:
> Because we increment the timearg pointer, shutdown(8) accepts
> stuff like
>
> shutdown ++10
>
> as strtonum(3) allows a single leading '+'.
>
> Not incrementing leaves the offset in a valid format for strtonum(3)
> and disallows stu
On Sat, 31 Mar 2018 12:53:52 -0500, Scott Cheloha wrote:
> Simpler.
OK millert@
- todd
On Sat, 31 Mar 2018 15:13:03 -0400, David Hill wrote:
> memcpy can be used on freshly allocated memory. Fill in the free size
> for it.
OK millert@
- todd
On Sat, 31 Mar 2018 15:15:34 -0400, David Hill wrote:
> Add the free size. (allocated in mfs_vfsops.c)
>
> mfsp = malloc(sizeof *mfsp, M_MFSNODE, M_WAITOK | M_ZERO);
> devvp->v_data = mfsp;
OK millert@
- todd
On Sat, 31 Mar 2018 13:41:00 -0600, "Theo de Raadt" wrote:
> Let's see. SIGINFO starts. Imagine if the buffering is small (I
> don't think it is). If it was small, there could be multiple writes
> to the sequence. SIGINT arrives. This will result in stderr output
> being a bit garbled. This
On Mon, 02 Apr 2018 12:34:27 +0200, Martin Pieuchot wrote:
> Here's a diff to move FREF() just after fd_getfile() in sys_kevent(),
> sys_lseek() and getvnode().
Looks fine. OK millert@
- todd
On Mon, 02 Apr 2018 12:51:06 +0200, Martin Pieuchot wrote:
> I'd like to ease reviews of the FREF()/FRELE() dances by keeping their
> number small. In the case of sys_fstatfs() it's easy to have a single
> FRELE() for the corresponding getvnode(). Ok?
I actually find the original easier to read
On Mon, 02 Apr 2018 10:40:15 -0500, Scott Cheloha wrote:
> What do you think about the signal race mentioned here:
>
> https://marc.info/?l=openbsd-tech&m=152260519904692&w=2
>
> deraadt pointed out that blocking SIGINT while SIGINFO is in summary()
> has the effect of making the process un-INTabl
On Mon, 02 Apr 2018 21:17:59 +0200, Ingo Schwarze wrote:
> * if (a <= b) for (i = a; i <= b;
>doas not make sense to me.
>If a > b, we have i > b right away, so the loop isn't entered.
> * a > b ||
>does not make sense either, for the same reason:
>The for loop itself takes care
On Sat, 07 Apr 2018 10:46:01 +0200, Anton Lindqvist wrote:
> Since patch(1) no longer invokes ed(1), pathnames.h can be removed.
> _PATH_TMP is still used inside patch.c but including paths.h is
> sufficient.
OK millert@
- todd
On Sat, 07 Apr 2018 21:33:24 +0200, Jeremie Courreges-Anglas wrote:
> At least ports/archivers/fuse-zip fails to build with clang 6, because
> of stricter diagnostics:
>
> -->8--
> c++ -c -O2 -pipe-I/usr/local/include main.cpp \
> -Ilib \
> -o main.o
> main.cpp:138:5: error: constant e
On Mon, 09 Apr 2018 11:56:46 -0500, Scott Cheloha wrote:
> Always nifty:
>
> $ video -f '%d%d%d'
> video: 31668978116843009-2139062144: No such file or directory
>
> So, use strlcpy and check for truncation.
>
> ok?
OK millert@
- todd
On Fri, 13 Apr 2018 11:11:04 -0500, Scott Cheloha wrote:
> So that $SECONDS advances uniformly, independent of the system clock.
Why are you including sys/time.h? For struct timespect you only
need time.h which is already included.
In general, you only need sys/time.h for struct timeval or for
On Fri, 13 Apr 2018 19:42:51 +0200, Jeremie Courreges-Anglas wrote:
> The diff uses timespecsub:
Ah, OK. That's fine then.
- todd
On Mon, 16 Apr 2018 10:19:40 +0200, Martin Pieuchot wrote:
> Diff below does FREF(9) earlier instead of incrementing `f_count' by hand.
>
> The error path is also updated to call FRELE(9) accordingly.
Wouldn't it be less error prone to simply add:
if (fp != NULL)
FRELE(fp
On Tue, 17 Apr 2018 09:50:49 +0200, Martin Pieuchot wrote:
> Sure, here's an updated diff. It also moves the FRELE(9) in the error
> loop down as suggested by visa@.
OK millert@
- todd
On Tue, 01 May 2018 13:35:54 -0600, "Theo de Raadt" wrote:
> > b) Their working space should be independent of each other. This
> > isn't hard, just splitting kd->argbuf into kd->argbuf and
> > kd->envbuf. Seems a bit saner.
> >
>
> I think (b) would be the better solution, this seems
On Thu, 03 May 2018 13:58:35 +0300, Vadim Zhukov wrote:
> Here is patch for libkvm that fixes a few memory handling problems.
> Most changes are mechanical, with some exceptions:
>
> 1. Most notable: this splits argv buffer into argv-specific one
> and environ-specific one. This makes ps -e
On Thu, 03 May 2018 17:59:39 +0200, Otto Moerbeek wrote:
> Yes, looks good from reading. But all te extra checks before calling
> free can go. That's idiom from a *long* time ago.
There is more cleanup that can be done in this code. For example,
the use of 0 instead of NULL. But that can be a s
On Mon, 07 May 2018 19:21:37 +0200, Alexander Bluhm wrote:
> When looping over mount points, the FOREACH SAVE macro is not save.
> The loop variable mp is protected by vfs_busy() so that it cannot
> be unmounted. But the next mount point nmp could be unmounted while
> VFS_SYNC() sleeps. As the l
On Tue, 15 May 2018 20:57:02 +0200, Marco Pfatschbacher wrote:
> I think it's time to remove this artifact from 1996.
Agreed. OK millert@
- todd
501 - 600 of 1551 matches
Mail list logo