On Thu, Aug 16, 2018 at 04:20:54PM -0400, Bryan Steele wrote:
> This adds unveil to pflogd(8)
>
> pflogd(8) is a special case, residing in /sbin, it's a static PIE. As
> such, I thought it might be worth experimenting with execpromises here.
> This allows re-exec after privdr
This adds unveil to pflogd(8)
pflogd(8) is a special case, residing in /sbin, it's a static PIE. As
such, I thought it might be worth experimenting with execpromises here.
This allows re-exec after privdrop, and removes chroot(2) in favour of
only unveil(2) and pledge(2). I've left the code there
In order to be able to safely append to existing log files, pflogd(8)
attempts to validate/or move invalid/broken pflog pcap files out of the
way on its own. I noticed that this is not compatible with unveil(2), as
pflogd(8) would need to be able to rename(2) files in /var/log to
On Thu, Aug 02, 2018 at 07:04:31PM +0200, Florian Obser wrote:
> I have been told that this is going to fall into snaps soon. If you
> are doing weird (or normal) things with ifconfig, please test.
>
> In particular if you use rulefile.
>
> Thanks!
>
> diff --git ifconfig.c ifconfig.c
> index
On Wed, Jul 25, 2018 at 10:33:48PM +0100, Stuart Henderson wrote:
> On 2018/07/25 22:57, Alexandre Ratchov wrote:
> > On Wed, Jul 25, 2018 at 05:10:05PM +0100, Stuart Henderson wrote:
> > > On 2018/07/25 15:26, Alexandre Ratchov wrote:
> > > > On Wed, Jul 25, 2018 at 11:56:04AM +0100, Stuart
On Mon, Jul 09, 2018 at 11:38:13AM +1000, Jonathan Gray wrote:
> On Sun, Jul 08, 2018 at 12:15:39PM -0700, Thomas Frohwein wrote:
> > Hi,
> >
> > It appears that HD Audio from AMD's generation Ryzen can't handle MSI.
> > This leads to the bug that I reported here:
> >
> >
On Wed, Jun 20, 2018 at 07:21:01PM +0200, Reyk Floeter wrote:
> Hi,
>
> when we develop code in OpenBSD, we have the choice of reviewing or
> sharing our patches privately between individual developers, on an
> internal list, or here on tech@.
>
>
> >From https://www.openbsd.org/mail.html:
>
>
eturn (xstrdup(ms.out));
> + return (ms.out);
> }
> return (NULL);
> }
>
>
>
> On Wed, Jun 13, 2018 at 11:00:58PM -0400, Bryan Steele wrote:
> > magic_test returns a xstrdup'd string, which was then being xstrdup'd
> > again without freeing the original copy
magic_test returns a xstrdup'd string, which was then being xstrdup'd
again without freeing the original copy (leaking memory).
casts added to avoid clang warning
warning: assigning to 'char *' from 'const char *' discards qualifiers
[-Wincompatible-pointer-types-discards-qualifiers]
Happens while booting my AMD desktop w/ radeon.
lock order reversal:
1st 0xff01d4b2be30 vmmaplk (>lock) @
/home/brynet/Projects/current/src/sys/uvm/uvm_map.c:4433
2nd 0xff01d415a0a0 inode (>i_lock) @
/home/brynet/Projects/current/src/sys/ufs/ufs/ufs_vnops.c:1559
lock order
On Wed, Jan 31, 2018 at 10:41:48PM +0100, Christian Weisgerber wrote:
> Alexei Malinin:
>
> > If the base system supported multi-booting in MBR would the community be
> > interested in it?
>
> It should be easy to adapt FreeBSD's boot0. I don't remember any
> demand for it, and now MBR is on
On Thu, Jan 25, 2018 at 07:00:43PM +0100, Martin Pieuchot wrote:
> Diff below adds support for dumping USB transfers via bpf(4), including
> the tcpdump(8) bits.
>
> I'd like special review of the new bpf_tap(9) function I'm introducing:
> - is there a way to pass the header as a different
On Wed, Dec 13, 2017 at 04:19:03PM -0500, Bryan Steele wrote:
> On Wed, Dec 13, 2017 at 09:56:32PM +0100, Matthieu Herrb wrote:
> > On Mon, Nov 27, 2017 at 08:22:30PM +1100, Jonathan Gray wrote:
> > > cd /usr/xenocara/lib
> > > ftp https://mesa.freedesktop.org/archive/m
17.1.6 though.
>
> I've also checked the it builds and that glxgears works (slowly as
> expected) on loongson.
> --
> Matthieu Herrb
>
Hi,
The last time the Mesa 17 update came up there was also an issue with
the newly introduced GLSL shader cache with pledge(2), which brok
On Sat, Dec 09, 2017 at 06:44:44PM +0100, Theo Buehler wrote:
> This was discussed in a smaller circle and has been in snaps for two
> days, but I'd like to show this to a wider audience.
>
> Theo asked me to make sure that all our shells print a prompt including
> the hostname by default. The
Finally, the last part..
Drop root privileges in the monitor process before we service any DNS
lookups for the parser process, and if we know beforehand that we never
will, chroot(2).
-Bryan.
diff -Nru tcpdump_orig/privsep.c tcpdump/privsep.c
--- usr.sbin/tcpdump/privsep.c Tue Sep 19 13:31:52
We can now pledge(2) "stdio bpf" when using tcpdump '-n' (don't convert
addresses to names), or when writing packet captures to stdout '-w -',
removing the "rpath" promise.
In the normal use case, the pledge(2) now becomes "stdio rpath dns bpf"
'rpath' is still needed to handle late opening of
Remove the internal privsep "getline" code, which was passing lines over
the socketpair, this was used previously to load the pf.os fingerprints
for the unpriv process, replaced by explicit descriptor passing.
-Bryan.
diff -Nru tcpdump_orig/privsep.c tcpdump/privsep.c
---
My last diff was too big, here's an attempt to split it up for easier
review. Please apply them in order.
This hoists opening pf.os(5) fingerprints '-o' from the 'RUN' state to
the 'FILTER' state, this will allow for a reduced pledge(2) at runtime
in the monitor process.
-Bryan.
Index:
On Thu, Sep 14, 2017 at 01:40:31AM -0400, Bryan Steele wrote:
> On Wed, Sep 13, 2017 at 09:53:09PM -0400, Bryan Steele wrote:
> > On Wed, Sep 13, 2017 at 08:58:28PM -0400, Bryan Steele wrote:
> > >
> > > 'rpath dns' for DNS lookups
> > >
> >
> > T
On Wed, Sep 13, 2017 at 05:48:04PM +0200, Alexander Bluhm wrote:
> Hi,
>
> syslogd has special code for reporting errors before it has been
> initialized. Then it tries to log to console. For every message
> it reopens the console with file descriptor passing from the privsep
> parent. Of
On Wed, Sep 13, 2017 at 09:53:09PM -0400, Bryan Steele wrote:
> On Wed, Sep 13, 2017 at 08:58:28PM -0400, Bryan Steele wrote:
> >
> > 'rpath dns' for DNS lookups
> >
>
> To clarify, "rpath" is not needed for DNS lookups, but currently for
> other address t
On Wed, Sep 13, 2017 at 08:58:28PM -0400, Bryan Steele wrote:
>
> 'rpath dns' for DNS lookups
>
To clarify, "rpath" is not needed for DNS lookups, but currently for
other address to name translations, i.e: getrpcbynumer(3) and
ether_ntohost(3).
OpenBSD's tcpdump(8) is separated into two processes: The packet parser
that will chroot(2) if possible, drop to an unprivileged user, and then
pledge(2) itself tightly as "stdio" before entering its main loop.
And the priviledged "monitor" process, which opens and fdpasses bpf(4)
and any
On Thu, Sep 07, 2017 at 12:27:18AM -0400, Bryan Steele wrote:
> Hi,
>
> This turned out easier then pflogd thanks to the existing privsep design
> work done by [otto@] and canacar@ many years ago. While tcpdump isn't a
> daemon in the traditional sense, it isn't so uncommon for
On Thu, Sep 07, 2017 at 01:16:18PM +0200, Alexander Bluhm wrote:
> On Thu, Sep 07, 2017 at 12:05:11PM +0200, Sebastian Benoit wrote:
> > The reason you see the message on the console on halt and reboot is, because
> > syslogd is gone at that point, so the message goes to the console instead.
>
>
On Thu, Sep 07, 2017 at 06:58:28AM +0200, Otto Moerbeek wrote:
> On Thu, Sep 07, 2017 at 12:27:18AM -0400, Bryan Steele wrote:
>
> > Hi,
> >
> > This turned out easier then pflogd thanks to the existing privsep design
> > work done by deraadt@ and canacar@ many
Hi,
This turned out easier then pflogd thanks to the existing privsep design
work done by deraadt@ and canacar@ many years ago. While tcpdump isn't a
daemon in the traditional sense, it isn't so uncommon for people to have
long running sessions. At least on OpenBSD, this is even safe thanks to
pflogd(8) currently spams the console on shutdown if syslogd wins the
race to die, this logging probably comes from the fact that pflogd was
largely based on syslogd.
Also, cleanup some missed SIGCHLD handling code that is no longer
neccesary, parent will exit when child closes its side of the
fork+exec model for pflogd(8); moves pcap init to the re-exec'd privsep
parent and uses 'legit' fdpassing primitives to send the bpf fd to the
unprivileged child process.
I've tried to keep this difff small so that it can go in, there's more
work to be done in the future, but this will let things
Hi,
vi(1) currently has support for filename tab expanding, but to expand the
tilde '~' character to $HOME it has to vfork/exec ksh. Yikes.
770 vi CALL execve(0x2cf0b7bce80,0x7f7bf960,0x2cef006cc00)
770 vi NAMI "/bin/ksh"
770 ksh RET execve 0
90346 vi RET
On Wed, Aug 16, 2017 at 11:09:57AM -0400, sven falempin wrote:
> 6.1 got a firmware (ewww) for seabios
>
> i mean this : /usr/ports/sysutils/firmware/vmm
>
> If i compile this ports on 6.0 do i have any chance it does something right
> or i am just digging my grave deeper ?
>
> Best,
>
>
> --
On Tue, Aug 01, 2017 at 05:56:54PM +0200, Stefan Fritsch wrote:
> Hi,
>
> we have the problem that on some HP laptops with i219V, it sometimes
> happens that em fails to attach with this error:
>
> em0: Hardware Initialization Failed
> em0: Unable to initialize the hardware
>
> It seems there
On Fri, Jul 28, 2017 at 12:09:17AM +0100, Marco Bonetti wrote:
> It looks like spaces and tabs got mangled, I'm resending the diff in-line in
> this email.
>
> Any other comments / OKs ?
>
> Index: etc/etc.amd64/sysctl.conf
> ===
>
On Tue, Jul 18, 2017 at 02:48:08AM +0200, Klemens Nanni wrote:
> Add missing verb and period.
>
> Index: driver/xf86-video-intel/man/intel.man
> ===
> RCS file: /cvs/xenocara/driver/xf86-video-intel/man/intel.man,v
> retrieving
On Mon, Jul 10, 2017 at 03:22:48PM +0200, Ingo Schwarze wrote:
> Hi,
>
> Bryan Steele wrote on Mon, Jul 10, 2017 at 08:21:19AM -0400:
>
> > Instead of relying on the window manager, why not add it to your
> > shell?
> >
> > https://deftly.net/posts/2017
On Mon, Jul 10, 2017 at 07:59:15AM +0100, Stuart Henderson wrote:
> On 2017/07/10 00:56, Todd T. Fries wrote:
> > What are our options to replace it with equivalent functionality should you
> > feel the need to march on without it?
>
> ssh `cut -d' ' -f1 .ssh/known_hosts | dmenu`
Instead of
On Fri, Jul 07, 2017 at 12:44:35PM -0400, Okan Demirmen wrote:
> Hi cwm users,
>
> In converting this to use getline(3) over fgetln(3), I'm asking the question
> if
> this (imho) mis-feature belongs in a window manager. I've hinted at wanting to
> remove it in the past but there was opposition.
On Thu, Jul 06, 2017 at 07:37:19AM -0600, Todd C. Miller wrote:
> Sorry, HP-UX actually has the same behavior as us. Here's what I
> have so far:
>
> 4.4BSD strtol() behavior:
> HP-UX
> NetBSD
> OpenBSD
I had discovered with awolk@ and Dragonfly also shares this
behaviour:
On Sat, Jul 01, 2017 at 08:48:18PM -0400, Bryan Steele wrote:
> On Sat, Jul 01, 2017 at 06:53:29PM -0400, Bryan Steele wrote:
> > On Sun, Jul 02, 2017 at 12:26:19AM +0200, Ingo Schwarze wrote:
> > > I don't see a need for two char * variables here, that's an artifact
> > &
On Sat, Jul 01, 2017 at 06:53:29PM -0400, Bryan Steele wrote:
> On Sun, Jul 02, 2017 at 12:26:19AM +0200, Ingo Schwarze wrote:
> > I don't see a need for two char * variables here, that's an artifact
> > of the typical fgetln(3) complications. Just
> >
> &g
On Sun, Jul 02, 2017 at 12:26:19AM +0200, Ingo Schwarze wrote:
> I don't see a need for two char * variables here, that's an artifact
> of the typical fgetln(3) complications. Just
>
> char*line = NULL;
> size_t linesize = 0;
> ssize_t slen;
>
> while ((slen =
e to 0
> (it is currently uninitialized) before calling getline().
>
> On Sat, Jul 01, 2017 at 04:35:39PM -0400, Bryan Steele wrote:
> > On Sat, Jul 01, 2017 at 09:27:47PM +0100, Nicholas Marriott wrote:
> > >
> > > What is the point of doing this? fgetln works f
er.
http://marc.info/?l=openbsd-cvs=147231601627483=2
No worries if you want to keep it.
>
>
> On Sat, Jul 01, 2017 at 02:47:25PM -0400, Bryan Steele wrote:
> > Seems to work, does this look right?
> >
> >
I think I lost this part in my larger diff, make sure fds are closed
after testing files.
-Bryan.
Index: file.c
===
RCS file: /cvs/src/usr.bin/file/file.c,v
retrieving revision 1.63
diff -u -p -u -r1.63 file.c
--- file.c 28 Jun
Seems to work, does this look right?
-Bryan.
Index: magic-load.c
===
RCS file: /cvs/src/usr.bin/file/magic-load.c,v
retrieving revision 1.25
diff -u -p -r1.25 magic-load.c
--- magic-load.c1 Jul 2017 14:34:29 - 1.25
magic_load() is only called in main and fcloses magicfp immediately,
this removes a redundant fclose.
Index: magic-load.c
===
RCS file: /cvs/src/usr.bin/file/magic-load.c,v
retrieving revision 1.24
diff -u -p -u -r1.24 magic-load.c
My latest commit broke support for stdin input, whoops.
ok?
Index: file.c
===
RCS file: /cvs/src/usr.bin/file/file.c,v
retrieving revision 1.62
diff -u -p -u -r1.62 file.c
--- file.c 28 Jun 2017 15:42:49 - 1.62
+++
On Tue, Jun 27, 2017 at 11:40:49PM +0100, Nicholas Marriott wrote:
> Thanks. Comments inline.
>
> > [..]
>
> One member per line in structs please. Also you could reduce the amount
> of space here now to one tab.
>
> > [..]
>
> magic_load (which parses the magic file) is now before pledge and
>
On Tue, Jun 27, 2017 at 01:20:59AM -0400, Bryan Steele wrote:
> On Tue, Jun 27, 2017 at 12:26:08AM -0400, Bryan Steele wrote:
> Some unintentional changes crept in, here's another diff..
Sorry, last diff broke width calculation.. 3rd times the charm.
Index: Ma
On Tue, Jun 27, 2017 at 12:26:08AM -0400, Bryan Steele wrote:
> OpenBSD's file(1) implementation was written by nicm@, first introduced
> in 5.8, the inital design included a privileged parent process which
> forked an unprivileged child which would handle potentially unsafe
>
OpenBSD's file(1) implementation was written by nicm@, first introduced
in 5.8, the inital design included a privileged parent process which
forked an unprivileged child which would handle potentially unsafe
file parsing.
It also had 'sandboxing' using systrace(4), which required complex
In ksh(1), ^T is documented to be bounded to transpose-chars while
in emacs mode, not "x_stuff". While here, remove these poorly
described functions as well..
-Bryan.
Index: emacs.c
===
RCS file: /cvs/src/bin/ksh/emacs.c,v
On Mon, Jun 19, 2017 at 09:22:57PM -0400, Todd Mortimer wrote:
> Hello tech,
>
> I have attached a patch that converts NOP padding from the assembler
> into INT3 padding on amd64. The idea is to remove potentially conveinent
> NOP sleds from programs and libraries, which makes it harder for an
>
On Tue, Jun 06, 2017 at 07:43:02PM +0200, Adam Wolk wrote:
> Hi tech@
>
> While reading htpasswd and htpasswd handling in httpd I noticed that both use
> different APIs to handle encrypting/decrypting the passwords.
>
> - htpasswd uses the bcrypt API
> - httpd uses the new crypt API
>
> The
On Wed, Sep 28, 2016 at 01:34:52PM -0400, Okan Demirmen wrote:
> Hi,
>
> Curious what the reaction might be if I removed the little geometry
> window in the top-left corner for mouse/pointer based window resizes.
>
> We currently print the x/y dimensions only for mouse based actions; we
> don't
On Thu, Sep 15, 2016 at 10:14:51AM -0400, Ted Unangst wrote:
> Florian Obser wrote:
> > Not everything listening on localhost port 53 is a recursive resolver.
> > nsd(8) per defaults listens on 0.0.0.0 and will respond with REFUSED for
> > almost every query. asr stops in that case and does not
On Wed, Jun 15, 2016 at 12:32:00PM -0400, Bryan Steele wrote:
> On Wed, Jun 15, 2016 at 11:23:53AM -0500, joshua stein wrote:
> > On Tue, 14 Jun 2016 at 07:22:14 +0200, Joerg Jung wrote:
> > > > We should not get hung up too much about the driver name. If it is
>
On Fri, Feb 19, 2016 at 09:41:59PM -0700, Jack J. Woehr wrote:
> Noth wrote:
> >Unfortunately the suspend to ram and hibernate options don't work
> They don't. Proprietary undocumented hardware. "Doctor, it hurts when I do
> this." "Don't do that."
That's not very helpful.
> Noth wrote:
>
On Fri, Nov 27, 2015 at 08:20:40PM -0700, Jack J. Woehr wrote:
> man xterm mentions the -tb switch (create a toolbar/menubar) which works on
> other platforms (e.g., Cygwin).
>
> The switch is absent in the OpenBSD (5.8) implementation of xterm, there is
> no toolbar.
>
> Not terribly troubled
On Fri, Nov 06, 2015 at 12:47:08PM -0500, Okan Demirmen wrote:
> For 2 reasons, I'd like to change M1 (mouse-button1) menu from
> displaying only hidden windows to showing all windows.
> - this is the only menu that differs between mouse and keyboard
> invocations (kbd one shows all)
On Mon, Jun 29, 2015 at 01:02:44AM +0300, li...@wrant.com wrote:
Do you have a patch that achieves the same goal (that is, the goal he
stated, not the one you're reading into) that is up to your standards?
One that reversed the submission of the proposed patch, correcting it
and this
On Sat, Mar 28, 2015 at 12:22:10AM -0400, Brian Callahan wrote:
Hi tech@ --
Noticed this yesterday reviewing brynet@'s vi diff.
We have the appropriate isblank(3) in ctype.h but the header
wasn't being included in all the files that used it so vi was
falling back to a #define in
On Wed, Mar 18, 2015 at 09:33:53AM -0600, Theo de Raadt wrote:
[I mistakenly sent the wrong diff, sorry, here is the right diff]
This is a further proposal for unification of the random number
generator functions in our tree. It is based upon my previous work,
but takes things a bit
On Mon, Mar 02, 2015 at 10:21:39AM +, Stuart Henderson wrote:
On 2015/03/01 18:19, Bryan Steele wrote:
It seems gethostbyname(3) fails if the answer returned is too big, for
example chat.freenode.net contains a lot of records that will easily
trigger this.
This was introduced
On Mon, Mar 02, 2015 at 09:06:36AM -0500, Bryan Steele wrote:
..
I've also sent mail to eric@ and gilles@, no reply yet.
Nevermind, filtering my mail too well. :-)
It seems gethostbyname(3) fails if the answer returned is too big, for
example chat.freenode.net contains a lot of records that will easily
trigger this.
This was introduced as part of the asr libc resolver rewrite.
$ ping chat.freenode.net
ping: unknown host: chat.freenode.net
$ host
This was committed to the upstream demos, but we include
glxinfo/glxgears in xenocara.
http://cgit.freedesktop.org/mesa/demos/patch/?id=999b6950c644266bb871e79438751bdba2fa2a08
Does this show useful output for anyone?
-Bryan.
Index: glxinfo.c
On Mon, Feb 02, 2015 at 01:06:34PM +1100, Jonathan Gray wrote:
On Sun, Feb 01, 2015 at 06:39:39PM -0500, Bryan Steele wrote:
This was committed to the upstream demos, but we include
glxinfo/glxgears in xenocara.
http://cgit.freedesktop.org/mesa/demos/patch/?id
On Mon, Feb 02, 2015 at 01:06:34PM +1100, Jonathan Gray wrote:
While that commit is not in a released version of mesa demos
yet I agree it is worth having.
I'd rather update all of glinxfo than just pull this commit though.
Here is a diff which does that and I've converted the man page
to
On Mon, Dec 15, 2014 at 07:36:41PM +, mar...@martinbrandenburg.com wrote:
Revision 1.178 of xenocara/app/cwm/conf.c made lock and term always show
at top of application menu. I don't see a way to remove this. Will this
be permanent behavior now?
-- Martin
I see this too, my brain simply
On Thu, Dec 11, 2014 at 09:52:46PM -0800, Eugene Yunak wrote:
Thank you. Are there any specific good libraries you know of?
--
The best the little guy can do is what
the little guy does right
LibreSSL :-)
-Bryan.
On Fri, Nov 14, 2014 at 11:51:14PM +0100, Kamil Rytarowski wrote:
In downstream (except FreeBSD and libbsd consumers) there is missing
strtonum(3). To enhance mg(1) and catch its bugs in general I need to
start with improvement of its portability to other unsupported systems
(as I'm a consumer
On Thu, Oct 09, 2014 at 02:52:50PM +1000, David Gwynne wrote:
this is the only use of select inside libc i could find. it appears
pretty straightforward to replace with poll.
this compiles, but i have no idea how to test it.
any suggestions? or oks?
Is rcmd(3) even used by anything
On Tue, Sep 09, 2014 at 07:20:09PM +0100, Mark Cave-Ayland wrote:
Hi all,
Following up from my posts at the beginning of the summer, I'm pleased to
announce that as of today, qemu-system-sparc64 built from QEMU git master
will successfully install OpenBSD from an .iso and boot back into it
On Wed, May 14, 2014 at 05:39:38AM +0900, SASANO Takayoshi wrote:
Hello,
Recently I bought Arduino board which uses Nanjing QinHeng (WinChipHead)
CH340T USB-UART bridge via eBay, and I found uchcom(4) did not work.
At misc@, other user reported similar problem. [1]
The uchcom(4) comes
It seems Microsoft has a document in an annoying format (DOCX) that
contains a list of their _OSI strings, so, let's pretend to be Windows 8
and Windows 8.1 if the firmware asks us. This could avoid buggy AML
paths on systems that don't ship with Windows 7 anymore.
When the AML interpretor was stripped from acpidump(8) people weren't
pointed toward a better alternative. This should make it more obvious
that Intel's tools are there to play with.
-Bryan.
Index: acpidump.8
===
RCS file:
On Thu, Mar 13, 2014 at 02:22:13PM -0400, Bryan Steele wrote:
..
.Sh SEE ALSO
.Xr mem 4
+.Xr ports 7
+.Xr packages 7
+.Xr pkg_add 8
.Sh HISTORY
The
.Nm
I noticed the lack of comma seperation, I'd fix that before commit.. ;)
On Fri, Sep 09, 2011 at 03:04:19AM -0400, Brynet wrote:
I think this was a feature, right? :-)
Here it is again, this time with the man page bits.
UDP port scanning doesn't make much sense, so prevent -u and -z from
being used at the same time.
ok?
-Bryan.
Index: nc.1
On Mon, Feb 06, 2012 at 04:47:45AM +, Mark Lumsden wrote:
There is a CAVEAT section in the man page that should also be
amended, I suspect.
Heh, whoops. :)
Although useless on the initaiting machine, is it of any use to
be able to scan a range of UDP ports, for diagnotic reasons, and
to
101 - 181 of 181 matches
Mail list logo
