Re: 6.5 on EdgeRouter Lite: 1 CPU offline?

[Tobias Ulmer]

On Thu, Apr 25, 2019 at 06:14:04PM +0200, Harald Dunkel wrote:
> There is a suspicious message
> 
>   dev/ksyms: Symbol table not valid.
normal

> 
> Next it seems that one CPU is offline somehow. ???
> 
>   chester# sysctl -a | grep -i cpu
>   kern.ccpu=1948
>   hw.ncpu=1
>   hw.cpuspeed=500
>   hw.ncpufound=2
>   hw.ncpuonline=1

https://ftp.openbsd.org/pub/OpenBSD/6.5/octeon/INSTALL.octeon

search for numcores

Re: Firefox (and SeaMonkey) automatically creates a Desktop folder in $HOME

[Tobias Ulmer]

On Wed, Feb 14, 2018 at 07:42:53AM -0700, James Anderson wrote:
> Xianwen,
> 
> If you create ~/.config/user-dirs.dirs with the following lines it will
> prevent Firefox from creating those folders:
> 
> XDG_DESKTOP_DIR="$HOME"
> XDG_DOCUMENTS_DIR="$HOME"
> XDG_DOWNLOAD_DIR="$HOME"

Needs to be "$HOME/" exactly:
https://cgit.freedesktop.org/xdg/xdg-user-dirs/tree/xdg-user-dir-lookup.c#n126

> 
> 
> Jim
> 
> On Wed, Feb 14, 2018 at 7:23 AM, Xianwen Chen  wrote:
> 
> > Dear OpenBSD users,
> >
> > I find that Firefox and SeaMonkey automatically create a Desktop folder in
> > $HOME directory. I do not use Desktop folder in fvwm and I do not want it.
> > You probably encountered the same problem. How can this behavior of Firefox
> > be disabled?
> >
> > Sincerely,
> > Xianwen
> >

Re: Industrial use of line printers, does/would your company/organization use them with our lpd?

[Tobias Ulmer]

On Wed, Feb 17, 2016 at 11:49:30AM -0600, Chris Bennett wrote:
> After reading up on printers in use, I discovered that there is
> significant use of line printers due to their very low cost of
> consumables, production of a very long lasting output, unlike
> laser/thermal/inkjet printers and high reliability.
> 
> Is anyone using these in a high volume output setting (not like a
> restaurant or other low volume)?
> 
> If not using, but would like to, what is broken, missing or otherwise
> wrong with our lpd/lpr system?

The only thing wrong with lpd is nobody tedu'ed it yet.

No really, it is outdated beyond rescue. If you want to write a new
print job queueing system, sure, have fun. Maybe you can come up with a
'cups' that doesn't suck?

> 
> I do see that lpc, lpq, lprm are dinosaurs and have to be made extinct
> and replaced with something more functional with more information output
> and better capabilities.


> 
> Thanks,
> Chris Bennett

Re: softraid(4)/bioctl(8) vs. non-512-byte sectors disks

[Tobias Ulmer]

On Thu, Oct 08, 2015 at 08:42:14AM -0400, Kenneth R Westerback wrote:
> ...

It works fine, I'm exercising 4K sr crypto with rsync every night.

Commit it pretty please :) The remaining bugs don't find themselves

Re: Ubiquiti EdgeRouter Lite

[Tobias Ulmer]

On Tue, Aug 18, 2015 at 10:33:37PM -0400, dominik...@openmailbox.org wrote:
 On 2015-08-18 11:42, Tobias Ulmer wrote:
 On Tue, Aug 18, 2015 at 01:59:49PM +, Jona Joachim wrote:
 On 2015-08-18, Ted Unangst t...@tedunangst.com wrote:
  Predrag Punosevac wrote:
  Dear All,
 
  I am contemplating buying a new machine which will act as a router/DNS
  caching server for my home network. Is anybody currently running OpenBSD
  on the Ubiquiti Networks EdgeRouter LITE in that capacity? I saw that in
  June 2015 USB support was added which allows installing to local disk on
  machine. Can anybody point me to a work in progress documentation diff
  for installing 5.8 octeon port.  I am reading right now
 
  Here are my notes, which are basic, but should be enough to get you 
  through if
  you're familiar with openbsd.
  http://www.tedunangst.com/flak/post/OpenBSD-on-ERL
 
 Thank you very much for the write-up! I'm looking into buying hardware
 to build a small OpenBSD home router and this looks interesting.
 You say that the machine will not be able to serve as an IPSEC gateway.
 Is that when you consider Gigabit ethernet or do you think that even a
 10 Mbit connection will require too much computational power to do
 IPSEC on this machine?
 You also mention the usb driver which is not so reliable. I don't see a
 USB port on the machine. Is this an internal bus? I would be interested
 to use it with hostapd with a usb wifi nic.
 
 There is only one internal port, and you need that for storage. The
 internal flash is not supported and it's not all that much anyway.
 
 I've not tested it, but have my doubts uboot will boot from a usb
 storage device with a hub in between.
 
 
 Kindly,
 Jona Joachim
 
 You can upgrade the internal usb storage capacity. Not all usb keys seems to
 work though. A couple functioning models can be found at the Gentoo
 MIPS/ERLite-3 page (https://wiki.gentoo.org/wiki/MIPS/ERLite-3).
 

That's because the usb timeout in uboot is pretty short. I'm using an
external sata disk (1TB), works fine when you power sequence them right.
Of course when the device draws its power from the internal bus, it
needs to initialize quickly in order to respond to uboot probing in
time.

 I bought a 16 GB one to use in my ERL with OpenBSD but I haven't tested it
 since I can't connect to the router's serial port. Seems it could be caused
 by my cheap usb-to-serial cable with an CH340/341 which returns me only
 garbage. I can type commands and see that they have an effect but all text
 is garbage.

Are you sure you're using the right baudrate? I've had no trouble with a
self-made cable and ftdi/pl2302 converters. The ERL doesn't seem
particularly sensitive. You might indeed have an adapter from the famous
One Hung Lo factory...
Do you not have some old box with a real serial port to test your
theory?

 
 Do you guys have any idea? Or do I just need to buy a better usb-to-serial
 converter with an FTDI processor like this one:
 http://www.usconverters.com/usb-serial-adapter-xs882
 
 Thanks for the help.
 
 Dom

Re: Ubiquiti EdgeRouter Lite

[Tobias Ulmer]

On Tue, Aug 18, 2015 at 01:59:49PM +, Jona Joachim wrote:
 On 2015-08-18, Ted Unangst t...@tedunangst.com wrote:
  Predrag Punosevac wrote:
  Dear All,
  
  I am contemplating buying a new machine which will act as a router/DNS
  caching server for my home network. Is anybody currently running OpenBSD
  on the Ubiquiti Networks EdgeRouter LITE in that capacity? I saw that in
  June 2015 USB support was added which allows installing to local disk on
  machine. Can anybody point me to a work in progress documentation diff
  for installing 5.8 octeon port.  I am reading right now
 
  Here are my notes, which are basic, but should be enough to get you through 
  if
  you're familiar with openbsd.
  http://www.tedunangst.com/flak/post/OpenBSD-on-ERL
 
 Thank you very much for the write-up! I'm looking into buying hardware
 to build a small OpenBSD home router and this looks interesting.
 You say that the machine will not be able to serve as an IPSEC gateway.
 Is that when you consider Gigabit ethernet or do you think that even a
 10 Mbit connection will require too much computational power to do
 IPSEC on this machine?
 You also mention the usb driver which is not so reliable. I don't see a
 USB port on the machine. Is this an internal bus? I would be interested
 to use it with hostapd with a usb wifi nic.

There is only one internal port, and you need that for storage. The
internal flash is not supported and it's not all that much anyway.

I've not tested it, but have my doubts uboot will boot from a usb
storage device with a hub in between.

 
 Kindly,
 Jona Joachim

Re: segmentation fault during package build

[Tobias Ulmer]

On Fri, Dec 05, 2014 at 10:48:36AM +0100, Riccardo Mottola wrote:
 Hi Tobias,
 
 what you write is frightening :)
 
 Tobias Ulmer wrote:
 On Fri, Dec 05, 2014 at 01:18:18AM +0100, Riccardo Mottola wrote:
 Tobias Ulmer wrote:
 full dmesg please
 Here it is:
 OpenBSD 5.6 (GENERIC) #94: Wed Aug 13 13:54:32 GMT 2014
 m...@credogne.gentiane.org:/usr/src/sys/arch/sparc/compile/GENERIC
 real mem = 166998016 (159MB)
 avail mem = 159440896 (152MB)
 mainbus0 at root: SUNW,SPARCstation-20
 cpu0 at mainbus0: TMS390Z50 v0 or TMS390Z55 @ 50 MHz, on-chip FPU
 cpu0: physical 20K instruction (64 b/l), 16K data (32 b/l) cache enabled
 This CPU module (Voyager iirc) has issues. It's stable most of the time,
 but small programs (like chmod, cat, touch, ...) crash in random places.
 Voyager? I don't know, checking:
 http://mbus.sunhelp.org/modules/index.htm
 
 I would identify it as SM50, the latter revision, with the large heatsink,
 not the round one.

I just had a look, mine are 501-2708

 I suppose you mean by unstable under OpenBSD? It was very stable under
 solaris 2.5.

Yes, under OpenBSD. I don't know about other OS.

 Would it crash the program always in the same place?

No.

 
 It is the first time I run OpenBSD on this machine, converting it from
 Solaris. It have a dual-HyperSparc module from Ross, but during OpenBSD
 install it apparently failed, becoming unreliable. I hope it is just a
 coincidence of the age and that OpenBSD doesn't fry modules :) After a
 while it is up I get back into OBP and at reboot I get a memory failure. I
 suppose the cache controller or the cache went bad.

Cooling issue? My SS10/SS20 work fine with a HM150S-512

 
 I still have a SM40 I think, for emergency. I will try that and see, after
 being certain that I can reproduce these error from reboot to reboot. Once I
 get the crash I can type make install and reproduce it in the same place,
 but a make clean will have it work and have crash ln in another package.
 
 
 Riccardo

Re: segmentation fault during package build

[Tobias Ulmer]

On Fri, Dec 05, 2014 at 01:18:18AM +0100, Riccardo Mottola wrote:
 Tobias Ulmer wrote:
 full dmesg please
 Here it is:
 OpenBSD 5.6 (GENERIC) #94: Wed Aug 13 13:54:32 GMT 2014
 m...@credogne.gentiane.org:/usr/src/sys/arch/sparc/compile/GENERIC
 real mem = 166998016 (159MB)
 avail mem = 159440896 (152MB)
 mainbus0 at root: SUNW,SPARCstation-20
 cpu0 at mainbus0: TMS390Z50 v0 or TMS390Z55 @ 50 MHz, on-chip FPU
 cpu0: physical 20K instruction (64 b/l), 16K data (32 b/l) cache enabled

This CPU module (Voyager iirc) has issues. It's stable most of the time,
but small programs (like chmod, cat, touch, ...) crash in random places.

Re: missing packages for SPARC

[Tobias Ulmer]

On Tue, Dec 02, 2014 at 10:35:43PM +0100, Riccardo Mottola wrote:
 Hi,
 
 I was pkg_add'ing some essential packages on a freshly installed SPARC
 machine. I noticed that several packages are missing. I thought it was the
 mirror, but they are missing on the master ftp too.
 I know that some packages might not build on sparc or do not have sense on
 that platform, however I was looking for pretty general stuff: libxmsl,
 libxslt or subversion.

It looks like sparc 5.6 package were built without the modf fix :(

http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/lib/libc/arch/sparc/gen/modf.S

There isn't much that doesn't require python as a build-depends
somewhere...

You're welcome to help out. There is an open issue with bash and
setjmp/longjmp (guessing) that breaks dbus (iirc). I've lost countless
hours and gave up on that.

 
 Is this a problem? or is it deliberate? Sebastian, I know you used to stress
 your SPARCs :)
 
 Thank you,
 Riccardo

Re: segmentation fault during package build

[Tobias Ulmer]

On Wed, Dec 03, 2014 at 09:38:17AM +0100, Riccardo Mottola wrote:
 Hi,
 
 I am running OpenBSD 5.6 on Sparc [1]
 
 Since I did not find several packages available, I got ports (5.6 tar.gz
 version), unpacked it and started building.
 
 
 While I attempt to install libxml I get, while installing bzip2 dependency:
 
 install -c -o root -g bin -m 555 bzgrep bzmore bzdiff
 /usr/ports/pobj/bzip2-1.0.6/fake-sparc/usr/local/bin
 install -c -o root -g bin -m 444 bzip2.1 bzgrep.1 bzmore.1 bzdiff.1
 /usr/ports/pobj/bzip2-1.0.6/fake-sparc/usr/local/man/man1
 Segmentation fault (core dumped)
 *** Error 139 in /usr/ports/pobj/bzip2-1.0.6/bzip2-1.0.6 (Makefile:105
 'install': @cd /usr/ports/pobj/bzip2-1.0.6/fake-sparc/usr/local/man/m...)
 *** Error 1 in /usr/ports/archivers/bzip2
 (/usr/ports/infrastructure/mk/bsd.port.mk:2807
 '/usr/ports/pobj/bzip2-1.0.6/fake-sparc/.fake_done')
 
 
 If I just type make install again, it happens again, thus I would exclude
 a memory issue which makes thins more random, but it repeats in the same
 place. Perhaps a bad generated binary or a function call causing problems?
 
 I wanted to look for the core file, but can't find it. Where could it be?
 
 Cheers,
 Riccardo
 
 [1] OpenBSD 5.6 (GENERIC) #94: Wed Aug 13 13:54:32 GMT 2014
 m...@credogne.gentiane.org:/usr/src/sys/arch/sparc/compile/GENERIC
 

full dmesg please

Re: RISC-V ?

[Tobias Ulmer]

On Sun, Nov 16, 2014 at 04:15:15AM +, Luiz Roberto dos Santos wrote:
 Hi,
 There's any effort yet to iniciate a port for RISC-V?

No

 
 Regards,
 L.

Re: Sun/Cassini Quad Gigabit Card Not Detected

[Tobias Ulmer]

In my experience, cas(4) is slow and not very stable on sparc64. I used
it in a Blade 150 firewall.

On Sun, Nov 09, 2014 at 11:31:37AM -0500, Jeff wrote:
 Hi,
 
 I've installed OpenBSD 5.6 (i386) on a dual processor XEON box which
 has a 4 port Sun (Sun# 501-6738-10) Gigabit NIC card.  dmesg doesn't
 have any indication that the card is installed. Booting Linux shows
 the card as Sun/Cassini which I believe should be handled by the cas
 driver.
 
 pcidump shows the following (I think that the NS Saturn is the NIC
 card):
 
 Domain /dev/pci0:
  0:0:0: Intel E7505 Host
  0:0:1: Intel E7505 Error Reporting
  0:1:0: Intel E7505 AGP
  0:2:0: Intel E7505 PCI
  0:2:1: Intel E7505 PCI
  0:30:0: Intel 82801BA Hub-to-PCI
  0:31:0: Intel 82801DB LPC
  0:31:1: Intel 82801DB IDE
  0:31:3: Intel 82801DB SMBus
  1:0:0: NVIDIA Riva TNT2
  2:28:0: Intel 82870P2 IOxAPIC
  2:29:0: Intel 82870P2 PCIX-PCIX
  2:30:0: Intel 82870P2 IOxAPIC
  2:31:0: Intel 82870P2 PCIX-PCIX
  3:1:0: Broadcom BCM5703X
  3:2:0: Intel unknown
  4:0:0: NS Saturn
  4:1:0: NS Saturn
  4:2:0: NS Saturn
  4:3:0: NS Saturn
  6:1:0: TI TSB43AB22 FireWire
 
 
 Any ideas on how to get OpenBSD to recognise this card?
 
 Thanks!
 Jeff
 Any ideas on how to get OpenBSD to recognise this card?
 
 Thanks!
 Jeff

Re: tmux mutt and f1

[Tobias Ulmer]

On Mon, Aug 25, 2014 at 10:42:57PM +0200, frantisek holop wrote:
 does anyone know of a way to make urxvt
 play together nicely with mutt (and tmux)
 regarding the f1 key?  it works in xterm...
 
 macro index,pager f1 shell-escapeless 
 /usr/local/share/doc/mutt/manual.txtenter help

Works in urxvt for me. You're probably using the wrong TERM/termName setting.
Should be rxvt-256color and screen inside tmux

 
 -f
 -- 
 on a scale of 1 to 10, 4 is about 7.

Re: ada95 : gnat 4.6 : compile time warnings

[Tobias Ulmer]

On Tue, Feb 04, 2014 at 04:36:22AM -0500, Mayuresh Kathe wrote:
 hello,
 
 i was fiddling around with ada95 under openbsd 5.4 using gnat-4.6.
 
 i created a sample program as below.
 
 with Ada.Text_IO;
 use  Ada.Text_IO;
 
 procedure Hello is
   -- no variables required
 begin
   Put_Line (Hello, world!);
 end Hello;
 
 
 the code compiled (not cleanly) and gave me an executable.
 the messages thrown up during compilation are below.
 
 egcc -c hello.adb
 gnatbind -x hello.ali
 gnatlink hello.ali
 /usr/local/lib/gcc/x86_64-unknown-openbsd5.4/4.6.4/adalib/libgnat.a(adaint.o)(.text+0x3f3):
  In function `__gnat_os_filename':
 /usr/obj/gcc-4.6.4/build-amd64/gcc/ada/rts/adaint.c:714: warning: strcpy() is 
 almost always misused, please use strlcpy()
 /usr/local/lib/gcc/x86_64-unknown-openbsd5.4/4.6.4/adalib/libgnat.a(cstreams.o)(.text+0x12c):
  In function `__gnat_full_name':
 /usr/obj/gcc-4.6.4/build-amd64/gcc/ada/rts/cstreams.c:237: warning: strcat() 
 is almost always misused, please use strlcat()
 /usr/local/lib/gcc/x86_64-unknown-openbsd5.4/4.6.4/adalib/libgnat.a(adaint.o)(.text+0x1c8):
  In function `__gnat_try_lock':
 /usr/obj/gcc-4.6.4/build-amd64/gcc/ada/rts/adaint.c:552: warning: sprintf() 
 is often misused, please use snprintf()

This is normal. gnatlib works on a wide variety of platforms and
there is no autoconf or anything, thus it's all very rough and basic C.
In addition, many buffers are allocated on the Ada side, and the buffer
size is not known in the C files due to a variety of very poor design
choices.

I've tried fixing this rats nest before, and got lost. I don't plan on
trying again. In part because it pisses me off that the company which
effectively owns the copyright to this part of gcc doesn't give two
shits about the free version or anything but their commercial
offerings.

Re: Request for Funding our Electricity

[Tobias Ulmer]

On Fri, Jan 17, 2014 at 08:10:02PM -0800, William Ahern wrote:
[...]

Compared to your suggestions, Die Hard 2-5 didn't contain any plot holes
and made perfect sense.

You are not arguing, but obviously, emulators are so much better.

With just a couple of modern Xeon machines (these are free, obviously),
writing or patching up a couple of emulators (with performant JIT backends,
of course), we could easily halve(!) our power bill.

How to emulate a bunch of +-1GHz SMP RISC machines or even just a 200MHz
sparc at native speed is, of course, left as a finger exercise to the
developers, while they also improve OpenBSD as per usual.

Just stop, please.

Re: Problem with dhcp requests on --current of Nov 2-4

[Tobias Ulmer]

On Mon, Nov 04, 2013 at 06:22:37PM -0500, STeve Andre' wrote:
 Sometime between Oct 18th and now I've lost the ability to do a
 dhcp request at boot time.  Dropping back to the Oct 18 kernel
 fixes things so I don't have hardware problems.
 
 When I attempt to do a dhclient em0 I get
 
 DHCPREQUEST on em0 to 255.255.255.255 port 67
 DHCPACK from 192.168.1.1 (00:1a:70:f8:07:38)
 
 and thats it.
 
 This is a thinkpad W500 running -current oct 18th that works, Nov 2-today
 kernels that do not work.
 
 Has anyone else seen this?

Yes, you want to read current.html and get a new userland.

 
 --STeve Andre'

Re: crypto softraid DUID's

[Tobias Ulmer]

On Thu, Oct 17, 2013 at 07:50:43AM +0200, frantisek holop wrote:
 hi there,
 
 if i have a usb key, that is softraid encrypted,
 it has 2 DUID's. the first one (before bioctl)
 can be used to script bioctl when the key is inserted.
 when the SR CRYPTO drive is attached, it has another
 DUID.  this can be used for mounting/unmounting.
 
 my question is, is that a security threat to have this
 2nd DUID in /etc/fstab?  could it be used as cleartext
 for brute forcing the SR CRYPTO drive?

There are many other known plaintext bits in a filesystem, starting with
the FFS magic marker. I don't think DUID matters at all in this context.
Now whether it's possible to mount a known plaintext attack on AES-XTS,
I leave that to the cryptographers.

 
 i also noticed that bioctl -c C -l accepts DUID's,
 but bioctl -d does not.  it this by design?

I would love a fix for this...

 
 -f
 -- 
 i'm weird, but i'm saving up to be eccentric.

Re: yacy on openbsd

[Tobias Ulmer]

On Sun, Aug 04, 2013 at 11:08:39AM +0300, Tony Berth wrote:
 Dear group,
 
 is anyone running yacy on a openbsd box?
 
 I tested the latest one yacy on a 5.3 amd64 but didn't succeed. The only
 resource I found was:
 
 http://ventejuy.es/cgi-bin/post?p=11051522005289 (in Spanish!)
 
 but was unable to connect to localhost:8090
 
 Thanks
 

I did play with it from time to time, and it worked fine. Haven't looked
at it recently. The main problem is that it's a ridiculous, ineffective
piece of Java bloatware that will suck the life out of any machine.

It's pretty sad, had they implemented it with something sane (and by
that I'm more talking about the style of software development, than the
language), I guess it would see quite some use...

Re: Non-intel desktop/laptop

[Tobias Ulmer]

On Mon, Jun 24, 2013 at 09:33:54PM +0100, Laurence Rochfort wrote:
 Hi all,
 
 I'm looking for advice on what the best bet for well supported
 non-intel hardware would be.  Doesn't have to be lightning fast, but
 being able to run a modern browser at reasonable rate is a must.
 
 My initial thoughts were either a Mac PowerBook G5 or Sun Ultra 25/45.
 I really like the HP C series workstations, but it seems support is a
 bit lacking.

It's all nice to wish going non-Intel, but the reality is that there are
very few people using Linux or BSD on anything that isn't mainstream. As
a result you will run into bugs, bugs, more bugs and then some more the
further you deviate.

If you want to fix things yourself, fine. If you depend on others to
help you along, or want to watch video... forget about it.

Do some research whether the architecture you want to use has the
applications you need. sparc64, to give an example, doesn't have a
modern firefox (i hear that may change soon, but I don't believe it
yet).

 
 Cheers,
 Laurence.

Re: Fandom and dangers of Free Software

[Tobias Ulmer]

Btw, if anyone feels the need to reply, don't include misc@. We are not
interested.

Re: mixing ports and non-ports programs

[Tobias Ulmer]

On Sun, Apr 14, 2013 at 11:47:51PM +0200, Jan Stary wrote:
 On Apr 14 11:03:45, alan01...@gmail.com wrote:
  On 4/14/13, Marc Espie es...@nerim.net wrote:
   On Sat, Apr 13, 2013 at 11:40:09PM -0400, Alan Corey wrote:
  
  The question is what's the best way to mix current stuff in
 
 Run current, and create updated ports
 for the versions you need to run.
 
  but OpenBSD 5.0 isn't that different from 5.2.
 
 Go fuck yourself.
 

Take your own advice, please. We are sick and tired of people insulting
others on this mailinglist because of the miss-guided idea that his
somehow helps OpenBSD.

It does not - it is actively damaging the project.

Re: Intel hyperthreading w/ Atom E6xx OpenBSD 5.2?

[Tobias Ulmer]

On Fri, Mar 08, 2013 at 11:07:25AM +0100, Janne Johansson wrote:
 2013/3/8 Andres Perera andre...@zoho.com
 
  wait wait ~
  can someone comment on this
  http://www.daemonology.net/hyperthreading-considered-harmful/ ?
  is it still in vogue?
 
 
 I run very few obsd systems where people get to sit and measure cache
 behaviour while I generate ssl/ssh keys, so even if it can be done, the
 window seems small.
 Disabling HT may be good for other reasons, so if you are scared of that,
 just turn it off in the BIOS and forget about the problem.
 
 HT was good in a short period to accustom stiff old people to the fact that
 SMP was coming. It's mostly irrelevant now, unless you spend ALL your cpu
 on distributed.net/BOINC or similar house warming projects where all cores
 are executing the same small code path all the time.

I don't believe it. Someone please compile a kernel with
time make -j$(sysctl -n hw.ncpu) with or and without HT and report
back... (and please, not on a space heater xeon 4 from 2003...)

 
 -- 
 May the most significant bit of your life be positive.

Re: Intel hyperthreading w/ Atom E6xx OpenBSD 5.2?

[Tobias Ulmer]

On Fri, Mar 08, 2013 at 11:44:12AM -0500, Ted Unangst wrote:
 On Fri, Mar 08, 2013 at 11:54, Tobias Ulmer wrote:
 
  I don't believe it. Someone please compile a kernel with
  time make -j$(sysctl -n hw.ncpu) with or and without HT and report
  back... (and please, not on a space heater xeon 4 from 2003...)
 
 If you are splitting a core, your process will run about 50% slower.
 (On Atom, anyway.) But you will be running more processes at once.
 
 On a machine with a single core, this is strictly a win. e.g.,
 Compiling a file takes 1 second.  Compiling two at the same time with
 HT takes 1.5 seconds.  Compiling two sequentially without HT takes 2
 seconds.  1.5  2.  qed. :)
 
 If you have two or more cores, you have to balance this fact with your
 ability to load all cores.
 

Did a little benchmarking:

hw.model=Intel(R) Core(TM) i5 CPU M 520 @ 2.40GHz

No HT, -j2:
2m2.25s real 3m14.05s user 0m23.09s system
2m2.58s real 3m15.89s user 0m22.23s system
2m4.67s real 3m18.11s user 0m23.43s system

No HT, -j1:
3m46.46s real 2m38.15s user 0m18.60s system
3m46.46s real 2m39.36s user 0m18.81s system
3m46.33s real 2m38.07s user 0m19.22s system

HT, -j1:
3m48.97s real 2m32.24s user 0m17.35s system
3m47.57s real 2m32.95s user 0m15.58s system
3m47.49s real 2m32.77s user 0m15.82s system

HT, -j4
1m43.51s real 5m20.49s user 0m47.84s system
1m45.53s real 5m26.49s user 0m47.84s system
1m47.28s real 5m31.17s user 0m48.16s system

I know which setting I would use.

zzz

Re: Xorg + wsudl(4) DL-165 consumes a lot of cpu

[Tobias Ulmer]

On Sat, Mar 02, 2013 at 06:05:27PM +0100, Alexis de BRUYN wrote:
 Hi Everybody,
 
 On OpenBSD 5.3/amd64 (full dmesg(8) at the end), while watching Youtube
 videos through Minitube, Xorg(1) and Minitube consume a lot of CPU (see
 top(1) below), and the video is lagging.
 
 My below xorg.conf is a 3-screen configuration with 3 USB2DVI devices
 (DisplayLink CONV-USB2DVI, with a DL-165 chip) and the wsudl(4) driver.
 
 I have also tested with one DL-195 only and I had the same result.
 
 All is working fine with the embedded graphic chip of the computer :
 vga1 at pci0 dev 2 function 0 Intel HD Graphics 3000 rev 0x09
 
 Is it a hardware or driver limitation, or misconfiguration ?

Your resolution is 1600*1200*3 bytes (lets ignore whatever protocol the
driver uses, probably not 3 byte words), so 1 full frame update, without
the large USB overhead, means pushing out 5.5MB/s. At just 25fps that is
138MB/s or around 1.2Gbit/s.

Now, USB2 has a theoretical bandwidth of 480Mbit/s, DVI has 4Gbit/s
(dedicated, with no protocol that also supports mice and stuff).  And
DVI gets replaced by DP (8Gbit/s and more) because it can't keep up.

Even if OpenBSD was super-awesome at graphics, video and usb (which it
clearly is not), it could not overcome the hardware limitations. Unless
the chip itself could do video decoding in hardware, but that is clearly
not something supported by a framebuffer driver.

So... no way is this ever going to work.

Your graphics card does nothing here, afaict. It's all purely done on
the CPU. Actually I find it kind of amazing that it works at all...

 
 Thanks a lot for your help.
 
 
 # top
 load averages:  3.34,  1.63, 
 0.71  
   
 test.temp.com 16:25:21
 35 processes: 34 idle, 1 on processor
 CPU0 states: 14.0% user,  0.0% nice, 13.6% system, 37.7% interrupt,
 34.7% idle
 CPU1 states:  0.0% user,  0.0% nice,  0.0% system,  0.0% interrupt, 
 100% idle
 CPU2 states:  0.0% user,  0.0% nice,  2.2% system,  0.0% interrupt,
 97.8% idle
 CPU3 states:  0.4% user,  0.0% nice,  1.4% system,  0.0% interrupt,
 98.2% idle
 CPU4 states:  0.8% user,  0.0% nice,  5.8% system,  0.0% interrupt,
 93.3% idle
 CPU5 states:  3.6% user,  0.0% nice, 33.3% system,  0.0% interrupt,
 63.1% idle
 CPU6 states:  7.8% user,  0.0% nice, 35.1% system,  0.0% interrupt,
 57.1% idle
 CPU7 states: 14.5% user,  0.0% nice, 37.2% system,  0.0% interrupt,
 48.3% idle
 Memory: Real: 109M/327M act/tot Free: 7547M Cache: 130M Swap: 0K/8197M
 
   PID USERNAME PRI NICE  SIZE   RES STATE WAIT  TIMECPU COMMAND
 29293 alexis 20  101M   78M sleep/3   poll  3:22 129.83%
 minitube
 12128 _x11   20   10M   31M sleep/3   select1:13 52.88% Xorg
 17482 _sndio 2  -20  480K  944K sleep/0   poll  0:13  3.32% sndiod
 11665 alexis 20  544K 1556K idle  select0:01  0.00%
 dbus-launch
 1 root  100  660K  420K idle  wait  0:01  0.00% init
 29037 _pflogd40  716K  376K sleep/5   bpf   0:01  0.00% pflogd
 22491 root  280 1188K 2556K onproc/6  - 0:01  0.00% top
  1717 root  180  760K 1756K idle  pause 0:01  0.00% xdm
 31119 root   20 3640K 3356K sleep/4   select0:01  0.00% sshd
  8338 alexis 20  976K 2892K idle  select0:01  0.00% fvwm
 21735 _syslogd   20  668K  840K sleep/4   poll  0:00  0.00% syslogd
 12279 root   20 1776K 1816K sleep/4   select0:00  0.00% sendmail
 16253 root   20 2204K 1332K idle  netio 0:00  0.00% Xorg
  6310 alexis 20  776K 1380K idle  poll  0:00  0.00%
 dbus-daemon
 14644 root  100 1396K 5348K idle  wait  0:00  0.00% xdm
 18751 root   20  752K 1420K idle  select0:00  0.00% sshd
 31544 root   20  652K  520K idle  netio 0:00  0.00% pflogd
  6195 _ntp   20  800K 1148K idle  poll  0:00  0.00% ntpd
 14903 alexis 20  564K 1944K idle  select0:00  0.00%
 FvwmPager
  3900 root   20  744K  980K idle  poll  0:00  0.00% ntpd
 16024 alexis 30 1200K 2332K idle  ttyin 0:00  0.00% bash
  9219 root   20  652K  848K idle  netio 0:00  0.00% syslogd
  3583 root   20  736K 1016K sleep/5   select0:00  0.00% cron
 10791 alexis 20 1888K 4468K idle  select0:00  0.00% xterm
 19389 root   20  504K  904K idle  netio 0:00  0.00% xconsole
  9006 _x11   20  632K 2804K idle  poll  0:00  0.00% xconsole
  7848 root   30  364K  972K idle  ttyin 0:00  0.00% getty
  7890 alexis180  584K  528K idle  pause 0:00  0.00% sh
  4067 root  100 1352K 2392K idle  wait  0:00  0.00% bash
 12627 _ntp   20  924K 1032K idle  poll  0:00  0.00% ntpd
 23233 root   30  456K  976K idle  ttyin 0:00  0.00% getty
 21623 root   30  508K  

Re: Are pthreads hopeless in 5.0?

[Tobias Ulmer]

On Sat, Jan 19, 2013 at 11:18:37PM -0500, Alan Corey wrote:
 They seem to work for my programs, but I'm trying to use rtl_fm from
 the osmocom group at http://sdr.osmocom.org/trac/wiki/rtl-sdr.  This
 machine boots into OpenBSD and an old Debian Linux and the program
 runs fine under Linux.  If I can figure out how to make it work under
 OpenBSD I'd like to submit some ifdefs for it.
 
 The program is for software defined radio using a $20 dongle plugged
 into a USB port that tunes 24 - 1700 MHz.  It works under Linux and
 Windows, but under OpenBSD the demodulation doesn't keep up with the
 incoming data. The driver uses a callback routine when it has data,
 which does a sem_post that should enable a demodulation thread.  It
 works after a fashion, but by the time it starts 5 or 6 more batches
 of data have come along and most of them get lost (in OpenBSD).  I
 think maybe I need to change the priority on the thread if possible.
 
 I posted this question originally to StackOverflow:
 http://stackoverflow.com/questions/14392158/pthread-priority-scheduling-under-openbsd
 
 Have I run into one of the shortcomings of pthreads as implemented in
 5.0?

Maybe, but 5.0 is not supported anymore (and the first suggestion would
by to try a new snapshot). We've switched to a completely different
threading model in 5.2.  See http://www.openbsd.org/52.html#new
Due to this, it makes even less sense to start looking into issues on
5.0.

 I'll probably upgrade to 5.3 when it comes out, but that
 involves hours of sitting in a car using a public WiFi connection
 which I don't plan to do until warmer weather.

Not sure what you're doing, but an update on a system that is properly
maintained takes probably 30 minutes, less if you have some practise.
You could also pull down the files and do the actual update at home.
It's more work the longer you wait...

 
   Alan
 
 -- 
 Credit is the root of all evil.  - AB1JX

Re: Small diff to calendar.birthday

[Tobias Ulmer]

On Wed, Jan 02, 2013 at 04:03:37PM -0700, Jeff Ross wrote:
 On 1/2/13 3:56 PM, mxb wrote:
 I think if you put District, the you should change Oblast to Province.
 
 You may know more than me--I took that directly from wikipedia :-)

Oblast should not be translated IMO. It's a loanword in many languages.
Then there is Krai, which is sort of the same thing. No - best leave it
alone.

I would however drop a village in , doesn't add anything useful.

 
 What should district be to be consistent and correct?
 
 
 
 //максим
 
 On 2 jan 2013, at 23:47, jr...@openvistas.net wrote:
 
 --- usr.bin/calendar/calendars/calendar.birthdaySun Oct 16 09:09:27 2011
 +++ usr.bin/calendar/calendars/calendar.birthday.newWed Jan  2 
 15:41:39 2013
 @@ -11,7 +11,7 @@
 01/01   J.D. Salinger born, 1919
 01/01   Heinz Zemanek born in Vienna, Austria, 1920
 01/01   Dolores Haze (prototype for Lolita) born, 1935
 -01/02  Isaac Asimov born in Petrovichi, Russian SFSR (now part of 
 USSR), 1920
 +01/02  Isaac Asimov born in Petrovichi, a village in Shumyachsky 
 District of Smolensk Oblast, Russia, 1920
 01/03   J.R.R. Tolkien born, 1892
 01/04   George Washington Carver born in Missouri, 1864
 01/04   Jakob Grimm born, 1785

Re: Arm Systems

[Tobias Ulmer]

On Mon, Dec 31, 2012 at 12:22:32PM -0800, Opie wrote:
 Hello,
 
 ? When looking to purchase an Arm system what should I be looking for?

Every single ARM SoC needs specific support by the OS, even if it's only
adding its unique Id. Attached hardware is (in general) not discoverable
like on other platforms.

From this it follows that you can only buy (supported) hardware that is
specifically listed on:

* http://www.openbsd.org/armish.html
* http://www.openbsd.org/zaurus.html
* http://www.openbsd.org/palm.html
* http://www.openbsd.org/beagle.html

Now, reality check: The hardware listed there is very old with the
exception of maybe beagle, some of efforts are dead and some developers
listed are not with the project anymore.

To put it kindly, ARM support is in very poor state regarding current
hardware. If you like get into kernel development, this might be a good
place to start**.

 ?I've seen a few with 1ghz cpu and 1gig mem but the gpu's are something I do
 not fully understand on how to look at it side by side with an ATI or Nvidia.
 ? I'm looking to have am arm system to use as a desktop, development, web
 serv, ftp, some media play in desktop.

For Linux such systems exists. For OpenBSD, I don't think so.

 ?It will be more for learning not just
 running all things on it at once. ?Any information would be helpful
 and?appreciated?very much!
 
 Thanks!
 Andy


** to the people who ask for tasks on misc@, here is one: Pick a board
(maybe your beloved rpi?) and start reading datasheets. Add support, how
hard can it be!

Happy New Year :D

Re: Running OpenBSD on Raspberry Pi

[Tobias Ulmer]

On Sun, Dec 30, 2012 at 05:01:23PM +0100, KarlOskar Rikås wrote:
 Hi, I wonder if it's possible to run OpenBSD on Raspberry Pi.
 
 Is there any image ready for putting on my SD card and boot up? If not, is
 there any manual or guide how to make one?

No it's not possible and there are no plans to change that. Search the
archives if you're interested in the reasons.

In short, there is plenty of better performing and better documented
hardware available for nearly the same price. This makes the rpi
unattractive for developers.

 
 Thanks.

Re: Running OpenBSD on Raspberry Pi

[Tobias Ulmer]

On Mon, Dec 31, 2012 at 12:28:51AM +, Heptas Torres wrote:
 Btw, is anyone working on porting OpenBSD on http://openpandora.org/ ?

It's somewhere at the bottom of my todo lists. Support should be mostly
there (in beagle).

Re: High performance IO (sendfile(), caching, and libev(ent))

[Tobias Ulmer]

On Thu, Dec 20, 2012 at 04:26:48AM -0500, Jean-Philippe Ouellet wrote:
 On 12/20/12 4:20 AM, Otto Moerbeek wrote:
  On Thu, Dec 20, 2012 at 04:06:52AM -0500, Jean-Philippe Ouellet wrote:
  
  On 12/20/12 3:53 AM, Jean-Philippe Ouellet wrote:
  and madvise() them to not be swapped out?
 
  Oops, I think I might have misinterpreted the meaning of MADV_WILLNEED.
  I think I meant mlock().
  
  Why trying to be smarter than the kernel? Mlocking pages will kill you
  if there's memory shortage.
  
  The kernel will try to keep much used pages in mem anyway.
  
  -Otto
  
 
 Okay, yeah. That's a terrible idea. But still, the question of direct
 file-to-socket sending vs. keeping copies in my address space and
 write()ing those to the socket still remains.

The file will be in the buffer cache. While it still takes a few
in-memory copies (which is what sendfile saves you), this should be fast
enough for most cases.

If you keep the data in your address space, you save one m-to-m copy,
but ignore all the benefits that the bc has compared to you (namely
knowing how much free memory really is available at runtime, not forcing
buffers into swap and more). You will probably end up shooting yourself
in the leg for a speed gain that probably can't be realized because the
network is the real bottleneck

Taking memory away from the kernel to duplicate functionality in
user-space is almost never a good idea.

 
 Normally I would just write both and profile them, but I can't figure
 out how to do the first on OpenBSD.

Re: issue tracker

[Tobias Ulmer]

On Wed, Dec 19, 2012 at 09:43:18PM +, sickm...@lavabit.com wrote:
 Hi,
 
 I have been using OpenBSD for quite a long time, and find it awesome.
 I've got some spare time lately and decided to hunt some bugs, but I
 don't really know where to start. Any suggestions?
 
 P.S. Yeah, I know about openbsd-bugs, but I suppose that's not all there
 is.
 

At the moment, bugs@ is all there is. I've not found it very helpful to
just stare at random bugs. If they are trivial, they usually get fixed
quickly. If they are hard, they are usually not solvable without serious
interest in the general area. And at that point, you don't need to ask
for tasks. There are (almost) no middle-ground bugs.

If you use OpenBSD, is there really no part that bothers *you*? Try to
identify and work on these first.

Re: Wireless WPA and crypto hardware

[Tobias Ulmer]

On Sat, Dec 15, 2012 at 09:52:28AM +0100, Martin Kjær Jørgensen wrote:
 Hi misc
 
 Can a wireless interface (say, athn) make use of a hardware crypto card
 like hifn when using WPA/WPA2 as encryption?

From a quick look through the kernel:

No, net80211 does not use the crypto framework, therefore it can not use
any hardware crypto devices.

Usually the chip itself can do the necessary operations, but that seems
disabled. It's all done in software, I think.

More generally, modern CPUs hardly break a sweat doing a few crypto
ops.  The overhead to talk to accelerator cards is usually large and
only makes sense on (really) slow hardware. Then there is the whole
issue of bus bandwidth and transferring the same data n times.

 
 /Martin

Re: AR9485WB-EG libre port

[Tobias Ulmer]

On Fri, Dec 14, 2012 at 10:12:48PM -0600, Maximo Pech wrote:
 Shut up and show us the code.

Some people have earned the right to reply like this, others have not.
Which one is it in your case?

Tobias

PS: Aren't you the guy who thinks PGP is essential in base, but can't
code?

Re: Wireless WPA and crypto hardware

[Tobias Ulmer]

On Sat, Dec 15, 2012 at 05:46:19PM +0100, Martin Kjær Jørgensen wrote:
 On Sat, Dec 15, 2012 at 11:03:00AM +0100, Tobias Ulmer wrote:
  On Sat, Dec 15, 2012 at 09:52:28AM +0100, Martin Kjær Jørgensen wrote:
   Hi misc
   
   Can a wireless interface (say, athn) make use of a hardware crypto card
   like hifn when using WPA/WPA2 as encryption?
  
  From a quick look through the kernel:
  
  No, net80211 does not use the crypto framework, therefore it can not use
  any hardware crypto devices.
  
  Usually the chip itself can do the necessary operations, but that seems
  disabled. It's all done in software, I think.
  
  More generally, modern CPUs hardly break a sweat doing a few crypto
  ops.  The overhead to talk to accelerator cards is usually large and
  only makes sense on (really) slow hardware. Then there is the whole
  issue of bus bandwidth and transferring the same data n times.
 
 Do you think an AMD Elan 133 Mhz is modern enough for at 54/mbit
 wireless WPA2 throughput?

No

 
  
   
   /Martin

Re: for students or your children

[Tobias Ulmer]

On Wed, Dec 12, 2012 at 11:03:50AM +0530, Jay Patel wrote:
 Hi all ..
 
 is OpenBSD taking part in google code-in :

The answer is No, as far as I'm aware. Did you have a specific project in
mind? You can get fame and glory without participating in Google-sponsored
events :)

 
 http://www.google-melange.com/gci/homepage/google/gci2012
 
 Thanks,
 Jay.

Re: xfsdump INTERRUPT

[Tobias Ulmer]

On Mon, Nov 19, 2012 at 02:10:09PM -0800, rlinsurf wrote:
 Hi--
 
 I'm trying to use xfsdump to copy all the files from my home DVR to a bigger
 hard drive.
 
 These are the instructions, with the old drive being sdb and the new drive
 being sda
 
 mkdir /mnt/fap
 mkdir /mnt/hr20
 mount -t xfs -o rtdev=/dev/sda3 /dev/sda2 /mnt/fap
 mount -t xfs -o rtdev=/dev/sdb3 /dev/sdb2 /mnt/fap
 
 and finally:
 
 xfsdump -J - /mnt/hr20 | xfsrestore -J - /mnt/fap
 
 At this point, I'm getting an error saying that I have a broken pipe (32)
 xfsdump interrupted 16 sec.
 xfsdump result INTERRUPT
 
 I've tried this sevral times, but now I'm getting to add -R, not saying to
 where, but even when I do add it to the xfsdump part of the command, it
 keeps saying broken pipe.
 
 What am I doing wrong?

Well, you seem to be asking Linux questions on an OpenBSD mailinglist.
We can't help with that, please post your questions to the correct list.

ps: I do like /mnt/fap. That's a nice and short name...

 
 Thanks.
 
 
 
 
 --
 View this message in context: 
 http://openbsd.7691.n7.nabble.com/xfsdump-INTERRUPT-tp219224.html
 Sent from the openbsd user - misc mailing list archive at Nabble.com.

Re: Building OpenConnect with libintl

[Tobias Ulmer]

On Thu, Nov 08, 2012 at 03:38:18PM +, Woodhouse, David wrote:
 On Thu, 2012-11-08 at 14:36 +0100, Marc Espie wrote:
 
  Pass LIBTOOL=/usr/bin/libtool on make's command line.
 
  Trying to get through the spaghetti of gnu autocrap only leads to
  insanity.
 
  That falls under the don't fight that shit, it's hopeless.
 
 Hm, OpenBSD libtool doesn't seem to honour the argument '-version-number
 2:1', so my resulting library is 'libopenconnect.so.0.0' instead of
 'libopenconnect.so.2.1'.
 
 Should I be specifying that differently?

Found this in the GNU libtool documentation: New projects should use
the -version-info flag instead.

It looks like -version-number is an accepted option (thus no error
message), but no further processing is done on it. May be an oversight
in our libtool?

Note that the ports system overrides the version specified on the
command line, to fit into the OpenBSD-specific versioning scheme.

 
 --
Sent with MeeGo's ActiveSync support.
 
 David WoodhouseOpen Source Technology Centre
 david.woodho...@intel.com  Intel Corporation
 
 [demime 1.01d removed an attachment of type application/x-pkcs7-signature 
 which had a name of smime.p7s]

Re: When to update -stable?

[Tobias Ulmer]

On Sun, Nov 04, 2012 at 12:05:22PM +, John Long wrote:
 I'm trying to remember how I should know when to update -stable. Is the
 errata web page the definitive source or is there some place else I should
 keep an eye on? 

sources-changes@ is the definitive source, you can filter for commits to
the stable branch if you don't care about the other stuff.

 
 Thanks,
 
 /jl
 
 -- 
 ASCII ribbon campaign ( ) Powered by Lemote Fuloong
  against HTML e-mail   X  Loongson MIPS and OpenBSD
and proprietary/ \http://www.mutt.org
  attachmentsCode Blue or Go Home!

Re: Kernel Level Audio Next Generation

[Tobias Ulmer]

On Tue, Jul 31, 2012 at 10:46:57PM +0300, Alexey Suslikov wrote:
 Hello misc.
 
 http://klang.eudyptula.org/

Heh, that's by the guy who got his ass whooped by Lennart at 27c3. His
talk made me cringe...

After watching, you may understand why he's writing his own stuff
instead of using the awesome PulseAudio.

 
 Just curious, why they didn't even try to evaluate OpenBSD sndio.

Because he's a Linux fanboi, isn't that obvious? :)

 
 An overall approach to the problem is interesting thing too
 
 Q: Why a audio system in the kernel?
 A: Because it's the only reasonable thing to do.
 
 What people think? Maybe we should write an article for wikipedia
 to make sndio more visible to rest of the world?

I think it's pretty pointless to document an audio system that's only
available on OpenBSD. Maybe add it to the main OpenBSD article if it
isn't mentioned there already.

 
 Regards,
 Alexey

Re: disk_map in subr_disk.c

[Tobias Ulmer]

On Wed, Jul 25, 2012 at 11:37:32PM +0200, Frank Brodbeck wrote:
 On Wed, Jul 25, 2012 at 10:51:56PM +0200, Otto Moerbeek wrote:
  Userland prorams do not share memory or symbols with the kernel at all
  that is a fundamental thing in Unix. Your code just references a bunch
  of uninitialized vars. 
  
  Chekc opendev(3) (source in src/lib/libutil/opendev.c) which is a
  userland function to do the translation you want. Note it interfaces
  with the kernel via ioctl(2), the actual work is done by the diskmap
  device driver that calls disk_map(), all in kernel mode.
 
 Funny, I first looked at opendev(3) and the DIOCMAP ioctl before I went
 on to disk_map(). Obviously I missed something. Will have a look at
 opendev(3) again. 
 
 Thanks a lot for the pointer.
 
 Frank.

Maybe this helps:

#include sys/types.h
#include sys/dkio.h
#include sys/disk.h
#include sys/ioctl.h

#include err.h
#include fcntl.h
#include limits.h
#include stdio.h
#include stdlib.h
#include string.h
#include unistd.h

int
main(int argc, char *argv[])
{
struct dk_diskmap dm;
const char *dmpath = /dev/diskmap;
char dev[PATH_MAX];
char *d;

if (argc != 2) {
fprintf(stderr, usage: duid2dev duid|device\n);
exit(1);
}

bzero(dm, sizeof(dm));
dm.flags = DM_OPENPART;

if ((dm.fd = open(dmpath, O_RDONLY))  0)
err(1, open: %s, dmpath);

strlcpy(dev, argv[1], PATH_MAX);
dm.device = dev;

if (ioctl(dm.fd, DIOCMAP, dm) == -1)
err(1, ioctl);

d = strrchr(dm.device, '/');
if (d[1] == 'r')
dm.device = d + 2;
else
dm.device = d + 1;
dm.device[strlen(dm.device) - 1] = '\0';

printf(%s\n, dm.device);
close(dm.fd);
exit(0);
}

 
 -- 
 Frank Brodbeck f...@guug.de

Re: load now over 1.00 all the time (i386, MP)

[Tobias Ulmer]

On Sat, Jun 30, 2012 at 10:01:53PM -0700, Philip Guenther wrote:
 On Sat, Jun 30, 2012 at 5:38 PM, frantisek holop min...@obiit.org wrote:
  it seems that since a couple of snapshots back,
  load never goes below 1.00 anymore on both of my
  notebooks (i386 MP).  what prompted me to write
  this email is that now my old thinkpad is affected
  as well.
 
  looking at top right after boot shows that load was normal
 
  load averages:  1.14,  0.85,  0.43
 
  but current load constantly being over 1.00,
  the averages eventually rise as well.
 
 Have you checked the output of 'top', perhaps with the -H option, and
 see what shows up as consuming CPU?

Here, have some load:
while sleep .9; do uptime; done

;)

 
 
 Philip Guenther

Re: macppc will it survive?

[Tobias Ulmer]

On Fri, Jun 22, 2012 at 10:37:09AM -0400, Nick Holland wrote:
 On 06/21/2012 01:27 PM, Peter J. Philipp wrote:
 Hi,
 
 Since deraadt mentioned the names of people who left to bitrig and I'm
 wondering what will happen to the macppc port?  Is it going to go the
 route of the mac68k port too?  I saw some commits earlier on it so that
 got my hopes up...
 
 I have a G4 Cube running OpenBSD/macppc and it has a lifetime of another
 2 years or so despite being 11 years old.  Its benefit is it's low watt
 draw (35 watts) and its silence no fans.  I replaced its hd with an ssd
 so it doesn't hum.  I sleep beside it iow.
 
 -peter
 
 I don't think OpenBSD/macppc is going away any time soon.
 I've never heard any mention of it...and if the talk about mac68k
 going away is any indication, the time between first talk and
 actually happening is something on the order of ten years :) (and
 again..I've seen no talk!).
 
 Mac68k basically went away because there was no one who cared to
 keep it up, and it really was a pain in the butt.  See Miod's post
 to mac68k@ for the details, but let me add one week for 'make
 build', plus another week for X, and that doesn't work anyway, and
 heck, it spends most of its time broke.  And that assumes the build
 worked.  Think about the frustration of running a build for four
 days...and getting an error.  You apply a fix, and ... four days
 later, you find something else. Yes, this happened to me recently.
 Think what this means to someone trying to bring a new feature to
 mac68k.

I doubt you could build mac68k in a week.

My HP 345 takes roughly two weeks to build src, if there are no
problems. IIRC 8-10h to build a kernel.

The interesting question really would be: Are there any plans to get rid
of m68k entirely. Because then I would have to switch to RusticBSD...

Another thing I'm wondering about is whether coldfire is compatible
enough to run the m68k userland.

 
 Plus...we have very little evidence anyone was actually USING mac68k.
 
 NONE of this applies to macppc.  The ONLY thing in common between
 mac68k and macppc in the OpenBSD project is the first three letters,
 and no one is confusing the two platforms.
 
 Nick.

Re: Large (3TB) HDD support

[Tobias Ulmer]

On Sun, Jun 03, 2012 at 01:09:16PM +0200, frantisek holop wrote:
 hmm, on Sat, Jun 02, 2012 at 10:44:22AM +, Christian Weisgerber said that
  Otto Moerbeek o...@drijf.net wrote:
  
I just fsck'ed a 2.7TB filesystem in 1 minute, 43 seconds.
61% full, 447166 files.

What CPU and how much RAM?  SATA2 or 3?
   
   Even more important: block size, fragment size, # of inodes?
  
  Default values all the way.  64k/8k.
  
  Filesystem SizeUsed   Avail Capacity iused   ifree  %iused  Mounted 
  on
  /dev/sd1d  2.7T1.6T1.0T61%  447167 91273535 0%   /export
  
  Watching this with top, I see fsck_ffs grow to a measly ~44 MB
  resident size.
 
 these must be some really nice disks :]
 
 for example only a 200G slice (also 64k/8k) of music/film/picture
 collection (not even full yet) on a notebook disk (5400 RPM) takes ages:
 
 Filesystem SizeUsed   Avail Capacity iused   ifree  %iused  Mounted on
 /dev/sd0d  217G153G   63.5G71%   44815 7197423 1%   /data
 
 $ time sudo fsck -f /dev/sd0d
 ** /dev/rsd0d
 ** File system is already clean
 ** Last Mounted on /data
 ** Phase 1 - Check Blocks and Sizes
 ** Phase 2 - Check Pathnames
 ** Phase 3 - Check Connectivity
 ** Phase 4 - Check Reference Counts
 ** Phase 5 - Check Cyl groups
 44815 files, 20076091 used, 8329340 free (13748 frags, 1039449 blocks, 0.0% 
 fragmentation)
 4m58.26s real 0m22.50s user 0m7.28s system

This comes down to the FFS1 vs FFS2 difference. Newfs will select FFS2
for bigger filesystems, reducing fsck times significantly at the expense
of more efficient disk space allocation in FFS1.

 
 
 i am more than curious about your amd thread, i am trying to get
 rid of fsck times by creative disklabeling and mounting read-only...
 
 -f
 -- 
 there are 3 kinds of people: those who can count  those who can't.

Re: bad 5.1 package of poptop

[Tobias Ulmer]

On Sat, May 26, 2012 at 02:31:52PM +0200, Federico Giannici wrote:
 The 5.1 package of poptop poptop-1.3.4p3.tgz (taken from 2
 different official FTP sites) seems to be compiled for a platform
 after 5.1.
 
 If I try to install it an error appears about a c library version
 not present:
 
 # pkg_add poptop-1.3.4p3.tgz
 Can't install poptop-1.3.4p3 because of libraries
 |library c.64.1 not found
 | /usr/lib/libc.so.50.1 (system): bad major
 | /usr/lib/libc.so.60.1 (system): bad major
 | /usr/lib/libc.so.51.0 (system): bad major
 | /usr/lib/libc.so.53.1 (system): bad major
 | /usr/lib/libc.so.56.0 (system): bad major
 | /usr/lib/libc.so.62.0 (system): bad major
 
 
 If I force the install, the ldd command confirm the problem:
 
 /usr/local/sbin# ldd ./pptpd
 ./pptpd:
 ./pptpd: can't load library 'libc.so.64.1'
 ./pptpd: exit status 4
 
 
 Thanks.
 

On which architecture?

I can observe no such problem:

argon:~$ echo $PKG_PATH
http://ftp.eu.openbsd.org/pub/OpenBSD/5.1/packages/i386/
argon:~$ sudo pkg_add -i poptop
poptop-1.3.4p3: ok
The following new rcscripts were installed: /etc/rc.d/pptpd
See rc.d(8) for details.
Look in /usr/local/share/doc/pkg-readmes for extra documentation.
argon:~$ ldd `which pptpd`
/usr/local/sbin/pptpd:
StartEnd  Type Open Ref GrpRef Name
1c00 3c004000 exe  10   0  /usr/local/sbin/pptpd
0b939000 2b967000 rlib 01   0  /usr/lib/libc.so.62.0
08d4 08d4 rtld 01   0  /usr/libexec/ld.so
argon:~$ dmesg|sed 2q
OpenBSD 5.1 (GENERIC) #160: Sun Feb 12 09:46:33 MST 2012
dera...@i386.openbsd.org:/usr/src/sys/arch/i386/compile/GENERIC
argon:~$

Re: bad 5.1 package of poptop

[Tobias Ulmer]

On Sat, May 26, 2012 at 04:22:41PM +0200, Federico Giannici wrote:
 Oops!
 I forgot to say it's amd64.

carbon:~$ ftp 
ftp://ftp.eu.openbsd.org/pub/OpenBSD/5.1/packages/amd64/poptop-1.3.4p3.tgz
Trying 77.238.36.56...
Connected to ftp-prod-srv04.it.su.se.
220 ftp-prod-srv04.it.su.se FTP server ready.
331 Please specify the password.
230 Login successful.
Remote system type is UNIX.
Using binary mode to transfer files.
200 Switching to Binary mode.
250 Directory successfully changed.
Retrieving pub/OpenBSD/5.1/packages/amd64/poptop-1.3.4p3.tgz
local: poptop-1.3.4p3.tgz remote: poptop-1.3.4p3.tgz
150 Opening BINARY mode data connection for poptop-1.3.4p3.tgz (34753 bytes).
100% 
|**|
 34753   00:00
226 Transfer complete.
34753 bytes received in 0.32 seconds (104.60 KB/s)
221 Goodbye.
carbon:~$ tar xzf poptop-1.3.4p3.tgz +CONTENTS
carbon:~$ grep wantlib +CONTENTS
@wantlib c.62.0
@wantlib util.11.2


There's no problem with the packages. You seem to be mixing snapshots
with 5.1 somehow, check your PKG_PATH etc.


 
 Thanks.
 
 
 
 On 05/26/12 16:06, Tobias Ulmer wrote:
 On Sat, May 26, 2012 at 02:31:52PM +0200, Federico Giannici wrote:
 The 5.1 package of poptop poptop-1.3.4p3.tgz (taken from 2
 different official FTP sites) seems to be compiled for a platform
 after 5.1.
 
 If I try to install it an error appears about a c library version
 not present:
 
 # pkg_add poptop-1.3.4p3.tgz
 Can't install poptop-1.3.4p3 because of libraries
 |library c.64.1 not found
 | /usr/lib/libc.so.50.1 (system): bad major
 | /usr/lib/libc.so.60.1 (system): bad major
 | /usr/lib/libc.so.51.0 (system): bad major
 | /usr/lib/libc.so.53.1 (system): bad major
 | /usr/lib/libc.so.56.0 (system): bad major
 | /usr/lib/libc.so.62.0 (system): bad major
 
 
 If I force the install, the ldd command confirm the problem:
 
 /usr/local/sbin# ldd ./pptpd
 ./pptpd:
 ./pptpd: can't load library 'libc.so.64.1'
 ./pptpd: exit status 4
 
 
 Thanks.
 
 
 On which architecture?
 
 I can observe no such problem:
 
 argon:~$ echo $PKG_PATH
 http://ftp.eu.openbsd.org/pub/OpenBSD/5.1/packages/i386/
 argon:~$ sudo pkg_add -i poptop
 poptop-1.3.4p3: ok
 The following new rcscripts were installed: /etc/rc.d/pptpd
 See rc.d(8) for details.
 Look in /usr/local/share/doc/pkg-readmes for extra documentation.
 argon:~$ ldd `which pptpd`
 /usr/local/sbin/pptpd:
  StartEnd  Type Open Ref GrpRef Name
  1c00 3c004000 exe  10   0  /usr/local/sbin/pptpd
  0b939000 2b967000 rlib 01   0  /usr/lib/libc.so.62.0
  08d4 08d4 rtld 01   0  /usr/libexec/ld.so
 argon:~$ dmesg|sed 2q
 OpenBSD 5.1 (GENERIC) #160: Sun Feb 12 09:46:33 MST 2012
  dera...@i386.openbsd.org:/usr/src/sys/arch/i386/compile/GENERIC
 argon:~$

Re: German Government claims to be able to break PGP and SSH

[Tobias Ulmer]

On Thu, May 24, 2012 at 02:26:43PM +0200, Stefan Wollny wrote:
 Hi there!
 
 What do you guys think about the reliability of the news (unfortunatelly
 in German only) on www.golem.de
 (http://www.golem.de/news/bundesregierung-deutsche-geheimdienste-koennen-pgp-
 entschluesseln-1205-92031.html) that the German government claims to be
 able to break PGP and SSH. The official answer to some MPs and the party
 Die Linke is here:
 http://www.andrej-hunko.de/start/download/doc_download/225-strategische-fernm
 eldeaufklaerung-durch-geheimdienste-des-bundes
 
 For the non-German speaking (found on page 3 of the official document):
 
 Question:
 3. Is the technique used also able to at least in part decode and/or
 analyze encrypted communication (e.g. by SSH of PGP)?
 
 Answer:
 Yes, the technique used is in principle able to do this, depending on
 the way and quality of the encryption. (Yepp - that's the complete
 answer!)
 
 Is this some sort of Governmental FUD by just NOT adding s.th. like if
 the password/passphrase is weak enough?

Why, do you think, did they add the word grundsdtzlich (in principle)
to their short answer?

Wouldn't they look completely stupid and a waste of money if they said:
No, we're not able to decrypt SSH or PGP traffic/communication. We are
incompetent clowns and all we can do is try to brute-force weak user
passwords and install trojans to grab the key.

That's not going to happen.

 
 STEFAN
 
 ---
 Mail: ste...@wollny.de
 Gnu PG-Key ID: 0x9C26F1D0

Re: SSH, root can repeat commands with up arrow, others cannot

[Tobias Ulmer]

On Sun, Mar 11, 2012 at 02:43:42PM -0500, Chris Bennett wrote:
 This started for me a while back.
 Login as root, I can repeat older commands with up down arrows.
 History command shows history.
 
 su -l otheruser
 
 Cannot use up down arrows to access history.
 History command shows correct history.

You most likely set EDITOR to something containing vi. ksh parses that
and switches to vi mode. IMO it's a disgusting feature, but that
appears to be just me.

set -o emacs
set +o vi

 
 Login remotely as otheruser.
 Same problem.
 
 Chris Bennett

Re: Installing OpenBSD on a TS-412 Turbo NAS

[Tobias Ulmer]

On Sat, Mar 03, 2012 at 04:30:14PM +0100, Michal Mazurek wrote:
 Has anybody successfully installed OpenBSD on a QNAP TS-412 Turbo NAS? I'm
 looking for a NAS that I can keep in my room, and would like to run OpenBSD.

No, OpenBSD does not support the Kirkwood SoC in that device.

 
 -- 
 Michal Mazurek

Re: looking for hardware recommendations, x86 or otherwise.

[Tobias Ulmer]

On Mon, Jan 30, 2012 at 05:13:25PM +0100, Tobias Ulmer wrote:
 [..]
 PS: I'm ready to change my opinion about Broadcom by 1800, for just a
 couple of PDF uploads on their website...

A stripped datasheet has been released:
http://www.raspberrypi.org/archives/615

Direct dl:
http://dmkenr5gtnd8f.cloudfront.net/wp-content/uploads/2012/02/BCM2835-ARM-Peripherals.pdf

It's only ~200 pages, but better than nothing.

Re: looking for hardware recommendations, x86 or otherwise.

[Tobias Ulmer]

On Mon, Jan 30, 2012 at 08:24:49AM -0500, Dewey Hylton wrote:
 - Original Message -
  From: Theo de Raadt dera...@cvs.openbsd.org
  To: Dewey Hylton dewey.hyl...@gmail.com
  Cc: misc@openbsd.org
  Sent: Sunday, January 29, 2012 6:32:21 PM
  Subject: Re: looking for hardware recommendations, x86 or otherwise.
  
   i'm hoping the raspberrypi will eventually be supported on openbsd
   (if the hardware proves to be stable, $35 sounds GREAT) but i don't
   have the skills to go there myself.
  
  Wow.   Dream on.  It is a mess of firmware.  You know nothing of our
  history?
 
 i know a bit of the history, sure. i know nothing of the raspberrypi 
 firmware, however. :)
 

Well, people are clearly too busy raving about the price to ask the
really hard questions.

Is there documentation for the components on the Broadcom SoC? Afaik,
there is _no_ documentation. We know it's supposed to be a weird
architecture where the graphic core initializes the CPU and loads the
loader from SD. We also know that the GPU requires a large blob in
Linux.

The Linux source is the only reference to the hardware. I have not looked
(is it available already?) at it yet, maybe the Broadcom developers
document it well. Or maybe they don't and just use magic numbers.

In the case of OpenBSD, source code is not considered documentation! If
all you have is the Linux source, you must implement it in pretty much
the same way as Linux does. Including the bugs. That's not only extra
hard, it's a huge waste of time, especially if there are better
alternatives:

There are tons of affordable ARM development boards out there whose
makers publish mostly complete data sheets, errata and so on. Texas
Instruments with the Beagleboard/Pandaboard is just one of them. (*)

So, who should you reward, Broadcom with its anti-opensource attitude
(while making decent amounts of money with billions of Linux routers) or
companies who invest into making documentation available?


PS: I'm ready to change my opinion about Broadcom by 1800, for just a
couple of PDF uploads on their website...


(*) Let me drive a nail in the RP coffin: It's an outdated (dead) ARM11
design, unlike what TI offers. Its only redeeming feature is the very
fast GPU, which is very unlikely to ever work under OpenBSD, even if
Broadcom would publish the SoC data sheet.

Re: vnconfig: /dev/rsvnd0c: Device not configured

[Tobias Ulmer]

It's in the fine manual:

http://www.openbsd.org/faq/upgrade50.html#vnd

Re: dual dvi with 2 monitors, 1 dvi is not detected

[Tobias Ulmer]

On Sat, Jan 07, 2012 at 10:02:22PM +0100, Didier Wiroth wrote:
 Hmm ... unfortunately, I'm not a programmer. I guess I'll have to buy
 another brand/card.
 I had a look at the ati driver manual
 http://www.openbsd.org/cgi-bin/man.cgi?query=atiapropos=0sektion=0manpath=OpenBSD+Currentarch=i386format=html.
 I checked many online shops, but I was not able to find a PCIe dual
 dvi card that is supported. The supported chips listed in the openbsd
 manuals are really old, and most available ati cards are radeon = HD
 5000. May be you have any recommandations?

You don't need dual DVI, most cards these days have DVI and HDMI
outputs, which are two names for the same standard. All you need is an
adapter.

Any low-end, passively cooled HD4xxx should do.

Re: audit openbsd.org!

[Tobias Ulmer]

You can help audit openbsd.org! Download the software development kit
here http://www.openbsd.org/ftp.html and follow these instructions:
http://www.openbsd.org/anoncvs.html#using

That should get you started.

On Tue, Sep 13, 2011 at 10:38:27PM -0700, lancebaynes87 wrote:
 Recently hacked sites..: 
 
 
 http://www.linuxfoundation.org/
 
 http://www.utorrent.com/
 
 http://kernel.org/
 
 http://www.linux.com/
 
 
 
 Has the OpenBSD site ever been hacked? Are there any good audit processes to 
 check the servers?
 
 OpenBSD team, please be aware, there are many script-kiddies OR pro's who 
 wants the most secure OS website :\
 
 I love OBSD, and I don't want it to fail like kernel.org:
 
 
 http://blackhats.com/infosuck/0x007c.png
 
 
 Please pay more attention than in normal case! :\
 
 /just an anxious OBSD user/
 
 Thanks and sry for this e-mail.

Re: ksh: bad number (with leading zeroes) should not work for 0 - 7

[Tobias Ulmer]

On Mon, Sep 12, 2011 at 04:45:28PM -0500, Abel Abraham Camarillo Ojeda wrote:
 Some of our shell scripts that work with dates and do something like:
 
 month=`date +%m`
 something  month=$((month-1))
 
 Suddenly started crashing on august... there seems to be a bug identifying
 not-numbers (numbers with leading zeroes) before '08' (eigth), how to 
 reproduce:
 
 $ for i in 0{0,1,2,3,4,5,6,7,8,9}; do a=$i; a=$((a-1)); echo $a; done
 -1
 0
 1
 2
 3
 4
 5
 6
 ksh: 08: bad number `08'


The string 08 can not be converted into a number for doing
calculations. The 0 prefix signifies an octal number and 08 is not valid in
the octal system. Bash will also tell you that this can not be done.

Fix your script, add the leading zero after you're done with the
calculation.

 $
 
 
 Thanks.

Re: Loongson -- is it actually encumbered now?

[Tobias Ulmer]

On Fri, Sep 09, 2011 at 12:32:25AM +0200, ropers wrote:
 I have for some time quite covetously looked at hardware for this:
 http://www.openbsd.org/loongson.html (and at the Lemote Yeeloong
 netbook in particular). But I could never really afford new kit, so I
 still haven't bought any loongson hardware. But I'm still thinking
 about it.
 
 The big draw for me was the reported complete open-source-ness and
 unencumbered-ness of the whole hardware platform.
 
 Now I'm reading at Wikipedia that the Chinese have supposedly caved
 and coughed up some protection money to one or the other US
 Intellectual Property (haha) shakedown scheme or entity:
 http://en.wikipedia.org/wiki/Loongson#MIPS_patent_issues
 
 So does this mean that this platform is now to be regarded as
 patent-encumbered and no longer completely free (libre)?
 (That would kind of ruin the big appeal for me.)
 
 Thoughts?
 
 regards,
 --ropers


Yeah, it's non-free and you don't even get all the VHDL files for the
chips. Now please go and continue your whining on the FSF or gNewSense
mailinglists, it has absolutly no place here.

Re: Trim-Slice / NVidia Tegra 2 support

[Tobias Ulmer]

On Tue, Aug 30, 2011 at 01:17:19PM +0200, Christer Solskogen wrote:
 Does this thing work with OpenBSD/armish (or beagle maybe)? It uses a
 dual-core ARM Cortex A9 CPU.

No. And the chances for it ever running OpenBSD are slim. Nvidia, in
their usual fashion, do not release documentation for the SoC and likely
never will, if their graphic chips are any indication.

 
 -- 
 chs,

Re: Package mirrors

[Tobias Ulmer]

On Tue, Aug 09, 2011 at 05:06:08PM -0700, Luis Useche wrote:
 Hi Guys,
 
 Is there something going on with the package mirrors? They are empty :S

Standard procedure during release preperation while -release packages
are being built for all architectures. It takes a long time on the
slower ones...

 
 Thanks,
 Luis.
 
 luis@meg:~
 $ ftp 
 ftp://openbsd.mirror.frontiernet.net/pub/OpenBSD/snapshots/packages/amd64/
 Connected to openbsd-mirror01.roch.ny.frontiernet.net.
 220- =
 220- = Welcome to openbsd.mirror.frontiernet.net =
 220- =
 220-
 220-   releases:
 220- * ftp://openbsd.mirror.frontiernet.net/pub/OpenBSD/
 220- * http://openbsd.mirror.frontiernet.net/pub/OpenBSD/
 220- * rsync://openbsd.mirror.frontiernet.net/OpenBSD/
 220-
 220-   anoncvs:
 220- * anon...@openbsd.mirror.frontiernet.net:/cvs
 220-
 220-   cvsweb:
 220- * http://cvsweb.openbsd.mirror.frontiernet.net/
 220-
 220-   ssh connection rate limiting enforced by pf(4).  don't be a jerk.
 220-
 220-   report problems to j...@frontiernet.net.
 220-
 220 openbsd-mirror01.roch.ny.frontiernet.net FTP server ready.
 331 Guest login ok, send your email address as password.
 230 Guest login ok, access restrictions apply.
 Remote system type is UNIX.
 Using binary mode to transfer files.
 200 Type set to I.
 250 CWD command successful.
 ftp ls
 150 Opening ASCII mode data connection for '/bin/ls'.
 226 Transfer complete.
 ftp ^D221 Goodbye.

Re: Can command-line options be specified in any place?

[Tobias Ulmer]

On Wed, Jun 22, 2011 at 03:48:59AM +0200, Benny Lofgren wrote:
 On 2011-06-22 03.03, vadi...@gmail.com wrote:
  Please continue to use Linux.
  That's ugly, useless and dangerous.
  
  Oops, looks like that was a holy war type of question. Sorry I did
  not want to start that.
  
  If you want Linux, use Linux.
  
  It's not that I want specifically Linux. I've just decided to look for
  a system that cat satisfy me from the usability point of view. I do
  not care if that will be Linux or *BSD or Solaris or whatever  else.
  The main idea was that the work with the system should be a pleasure,
  not a pain :)
 
 What you should do is relearn the proper way. :-)
 
 Consider the fact that Unix have been around since the 1970's, and the
 *BSD flavor is as direct a descendant of the original look, feel and
 intent as you can possibly find today.
 
 Linux is, in that regard, an abomination. It's the bastard child of
 someone not properly trained in the unix way, who made stuff up
 as he went without regard for history, continuity, elegance or, for
 that matter, backwards compatibility.
 
 I feel the same way as you do, only the other way around. I really
 can't stand using a linux system for any length of time. Everything
 is similar, but different. Or different, but similar. And so darn stupid!
 
 Linus didn't do his homework properly. That, combined with the fact that
 Linux became such a huge success is both a blessing and a curse to us
 in the unix community; on the one hand Linux provides us with plenty of
 young blood in a new generation of hackers... while on the other hand
 they can't speak properly!

Oh please, Linus wrote the kernel, not Ubuntu. If you hate coreutils or
getopt, blame the respective groups that developed them and not someone
writing a kernel, a long time ago.

This rose tinted OpenBSD is the greatest shit really gets on my
nerves. It's all fun to bash others, but from time to time you have to
look at their stuff and figure out which parts they did right and you
could improve.

Blah blah misc@

 
 It's as if they've accidentally gone to veterinary school instead of
 medical school, without knowing it. Sure, they'd know just as much
 about anatomy as a real doctor would, but take my advice: if you're
 not a horse, don't go there for your pains...
 
 
 Regards,
 /Benny
 
 -- 
 internetlabbet.se / work:   +46 8 551 124 80  / Words must
 Benny LC6fgren/  mobile: +46 70 718 11 90 /   be weighed,
 /   fax:+46 8 551 124 89/not counted.
/email:  benny -at- internetlabbet.se

Re: Can command-line options be specified in any place?

[Tobias Ulmer]

tl;dr: In my opinion, these anti Linux rants do harm to OpenBSD by
condemning everything Linux does instead of allowing us to pick out just
the good parts.

On Wed, Jun 22, 2011 at 11:22:02AM +0200, Benny Lofgren wrote:
 On 2011-06-22 09.24, Tobias Ulmer wrote:
  On Wed, Jun 22, 2011 at 03:48:59AM +0200, Benny Lofgren wrote:
  On 2011-06-22 03.03, vadi...@gmail.com wrote:
  Please continue to use Linux.
  That's ugly, useless and dangerous.
 
  Oops, looks like that was a holy war type of question. Sorry I did
  not want to start that.
 
  If you want Linux, use Linux.
 
  It's not that I want specifically Linux. I've just decided to look for
  a system that cat satisfy me from the usability point of view. I do
  not care if that will be Linux or *BSD or Solaris or whatever  else.
  The main idea was that the work with the system should be a pleasure,
  not a pain :)
 
  What you should do is relearn the proper way. :-)
 
 
 [the rest of my rant deleted]
 
  Oh please, Linus wrote the kernel, not Ubuntu. If you hate coreutils or
  getopt, blame the respective groups that developed them and not someone
  writing a kernel, a long time ago.
 
 No, I don't hate coreutils or getopt, getopt is good shit. What I hate
 is the inconsistensies, the fact that Linux isn't a homogenous piece of
 work but so obviously a product of a thousand chefs, few with similar
 taste.
 
 And my criticism extends to the kernel too, or rather begins with it, so
 it definitley applies to Linus himself and the kernel guys.
 
  This rose tinted OpenBSD is the greatest shit really gets on my
  nerves. It's all fun to bash others, but from time to time you have to
  look at their stuff and figure out which parts they did right and you
  could improve.
 
 Granted, my rant was, on purpose, negatively Linux-biased, but not in
 one single place - also on purpose - would you have found the word
 OpenBSD or any slant towards it, which makes me suspect you couldn't
 stand what I wrote long enough to actually read all of it. :-)

Right. And I felt in the mood to take the opposite position for the fun
of it.

 
 So I think you might have missed my point. There is a true unix
 heritage that needs to be cared for, THAT MAKES LIFE SIMPLER if you
 understand and take advantage of it.

OpenBSD specifically and old BSD in general is not true to Unix. From
ksh to billions of options to find and other tools to the entire
networking framework (bolted on with additional syscalls, pseudo devices
etc), nothing of that is Unix (or even -like).

Here is something to read: http://harmful.cat-v.org/cat-v/

BSD went through a similar phase as GNU: adding every feature known to
man to the original Unix commands. Have a look at lpr(1) for GNUism in
action. After some time we got a little wiser and stopped adding flags
for everything that was convenient.

Linux, especially with the constant influx of new developers and
commercial interests, hasn't yet cooled down enough to stop messing
around with their base system.

However if I got my history right, the improvements of BSD are why people
bought a Unix license and then installed BSD. It was better, it had more
features, networking, usable error messages, better language support
etc. etc.

 Linus missed or chose to ignore that part entirely. That's fine, as
 Linux is not said to be a unix operating system, but a unix like one.
 
 The problem is, this likeness is not like enough, so it really
 doesn't help the community overall but rather hinders it. This is
 something the Linux and GNU folks could have addressed in the early
 days but either chose to ignore or were ignorant about. For that
 they absolutely deserve some blame.

Are you ready to test my patch where I'm going to remove -exec from
find(1) so you can have your real Unix back? And -r from grep? And...

Bullshit, you use BSD because just like Linux, it added lots of handy
features while keeping it simple. Linux may overdo it from your and my
point of view, but so does OpenBSD from the POV of some old unix guys.

 
 Now, the OP:s questions are certainly addressable by choosing a shell
 he is used to, and perhaps by a set of aliases and/or scripts to tune
 the user experience into something familiar for him.

The getopt(3) function is inconsistent amongst operating systems and
could use some polish in my opinion. Maybe there are technical reasons
why this feature can't be implemented, but this discussion has certainly
extinguished my curiosity about it.

Backwards threads like this one prevent people from trying to improve
things, which is the real damage done.

Once they get discussed in this manner on misc@, it's  difficult to
get even very sensible patches committed. Some developers may have
formed a strong anti stance and it takes years to convince them.

 
 My problem with that, and the reason for the recommendation I made
 before digressing into rant mode, is that that practice will get him
 into trouble in the long run, as he encounters other flavors

Re: Light being shone on GPU security

[Tobias Ulmer]

On Fri, Jun 03, 2011 at 01:29:40PM +, Kevin Chadwick wrote:
 http://www.contextis.com/resources/blog/webgl/;
 
 however this still pushes much of the responsibility of securing WebGL
 on the hardware manufacturers. Perhaps the best approach would be to
 design a specification for 3D graphics from the ground up with these
 issues in mind.
 
 Well many are going to get hurt in the process with having hw accel
 enabled by default in web browsers but atleast it may force improvements
 in security by the GPU manufacturers. With a #bit# lot of luck Nvidia
 may be pressurised to Open Source too.

That's going to be a very cold day in hell...

 
 Or is it just making browser code or even OS code for that matter, less
 secure?

Re: Wildest Africa Tour

[Tobias Ulmer]

On Tue, Apr 05, 2011 at 05:15:53PM +0200, Fridiric Perrin wrote:
 Stuart VanZee stua...@datalinesys.com writes:
  Don't be silly.  While Lions do provide excelent physical security
  they don't provide any data security at all.
 
 Just imagine for a moment protecting your OpenBSD boxen with a pair of
 lions --now we got real security! Much more than with a pufferfish
 that's going to gasp for air as soon as it is out of the water...

Lions are useless: http://www.youtube.com/watch?v=TBpu4DAvwI8

 
 -- 
 FrC)dC)ric Perrin -- http://tar-jx.bz

Re: OpenBSD 4.8 HP ProLiant DL360 G7 Install OK?

[Tobias Ulmer]

On Wed, Mar 09, 2011 at 08:03:25AM +0100, rancor wrote:
 We are running stock OpenBSD 4.8 on several DL360 G6 and G7 and a
 additional HP NC364T PCI Express Quad Port Gigabit Server Adapter. We
 are running them in pairs as a firewall cluster with pfsync and carp
 and also as stand alone as a router but we have problem with random
 freezes. We have tried both AMD64 and i386, all patched up to latest
 errata but we can't make them work stable. Now are we using VMware
 ESXi as a layer between the hardware and OpenBSD and every thing works
 just perfect, not a single freeze since we did this .. work around

 Our bug report on this matter:
 http://cvs.openbsd.org/cgi-bin/query-pr-wrapper?full=yesnumbers=6321

That bug report is not very useful to a developer and way too old. Get a
serial cable and run the latest snapshot (without disabling any
devices).  Capture dmesg and panic, type trace, ps - show
registers may also be helpful. No screenshots.

If you want help, YOU need to provide this information OR ship the hw to
someone who wants to look into it. Whining on misc@ is certainly not
going to fix anything.


 Since we don't know how to resolve this our future with OpenBSD is
 most unsure, we have more then 50 installations on different platforms
 and it's only G6 och G7 that are causing any problem.

 It's hard to say if it's only us since I never hear anything about
 this issue but we sure feels alone.

 Regards

 2011/3/9 ll
j7  sg...@ktis.kr:
  I am looking to purchase the server for OpenBSD 4.8
 
  OpenBSD 4.8 HP ProLiant DL360 G7 Install OK?
 
  Or another Recent HP Server ?
 
 
 
  Help me PLZ

Re: OpenBSD 4.8 HP ProLiant DL360 G7 Install OK?

[Tobias Ulmer]

On Wed, Mar 09, 2011 at 11:39:16AM +0100, rancor wrote:
 I have been asking for how I can provide more information but the question
 was not answered. Im not whining, Im just providing my experience.
 
 I would be happy to provide more information if some body asked my in my
 report and also told me what to do because I don't know and If I did I had
 provided that information in the first place.
 
 If you got better experience with DL360 G7 (and G6) please tell us how your
 setup was so that this user can install without any problem and I can
 resolve my issues. In other case please shut up with your lame attitude and
 start to help instead of beeing an arsehole.

Awww, how about you take a cold shower and read the mail again. The first
paragraph has all the steps to write a more detailed bug report.

 
 Regards
 Den 9 mar 2011 11.21 skrev Tobias Ulmer tobi...@tmux.org:
  On Wed, Mar 09, 2011 at 08:03:25AM +0100, rancor wrote:
  We are running stock OpenBSD 4.8 on several DL360 G6 and G7 and a
  additional HP NC364T PCI Express Quad Port Gigabit Server Adapter. We
  are running them in pairs as a firewall cluster with pfsync and carp
  and also as stand alone as a router but we have problem with random
  freezes. We have tried both AMD64 and i386, all patched up to latest
  errata but we can't make them work stable. Now are we using VMware
  ESXi as a layer between the hardware and OpenBSD and every thing works
  just perfect, not a single freeze since we did this .. work around
 
  Our bug report on this matter:
  http://cvs.openbsd.org/cgi-bin/query-pr-wrapper?full=yesnumbers=6321
 
  That bug report is not very useful to a developer and way too old. Get a
  serial cable and run the latest snapshot (without disabling any
  devices). Capture dmesg and panic, type trace, ps - show
  registers may also be helpful. No screenshots.
 
  If you want help, YOU need to provide this information OR ship the hw to
  someone who wants to look into it. Whining on misc@ is certainly not
  going to fix anything.
 
 
  Since we don't know how to resolve this our future with OpenBSD is
  most unsure, we have more then 50 installations on different platforms
  and it's only G6 och G7 that are causing any problem.
 
  It's hard to say if it's only us since I never hear anything about
  this issue but we sure feels alone.
 
  Regards
 
  2011/3/9 l l j7 sg...@ktis.kr:
   I am looking to purchase the server for OpenBSD 4.8
  
   OpenBSD 4.8 HP ProLiant DL360 G7 Install OK?
  
   Or another Recent HP Server ?
  
  
  
   Help me PLZ

Re: Radeon HD 4850 and drmRadeonCmdBuffer: -22.

[Tobias Ulmer]

On Mon, Feb 28, 2011 at 07:38:27PM +0100, Pascal Stumpf wrote:
 On Thu, Feb 24, 2011 at 10:03:06AM -0500, Joe Snikeris wrote:
  On Thu, Feb 24, 2011 at 3:09 AM, Andy Bradford
  amb-open...@bradfords.org wrote:
  
   Is it going to be possible to get acceleration working with this? If so,
   any pointers?
  
  +1
  
 
 For the record: I've also tried the radeonhd driver instead of radeon.
 'glxinfo' reports a lot more GLX Visuals and GLXFBConfigs, OpenGL
 version 2.1 instead of 1.5, but glxgears does not seem any faster
 (around 290 FPS). Other applications using GL (supertuxkart ...) still
 crash when trying to use hardware acceleration.
 
 For my part, I'd greatly appreciate *any* feedback from graphics people
 on this. Is it being worked on? Which component is the culprit (kernel,
 xorg driver, Mesa)?
 

There is no DRM support for r600 and up in the kernel. Someone first has
to finish his PhD first I believe...
With the current kernel support you get _some_ benefits like xvideo, but
that's about it for the moment.

Re: Radeon HD 4850 and drmRadeonCmdBuffer: -22.

[Tobias Ulmer]

On Mon, Feb 28, 2011 at 06:52:43PM -0500, Greg Jones wrote:
 On 02/28/11 15:16, Tobias Ulmer wrote:
 On Mon, Feb 28, 2011 at 07:38:27PM +0100, Pascal Stumpf wrote:
 On Thu, Feb 24, 2011 at 10:03:06AM -0500, Joe Snikeris wrote:
 On Thu, Feb 24, 2011 at 3:09 AM, Andy Bradford
 amb-open...@bradfords.org  wrote:
 Is it going to be possible to get acceleration working with this? If so,
 any pointers?
 +1
 
 For the record: I've also tried the radeonhd driver instead of radeon.
 'glxinfo' reports a lot more GLX Visuals and GLXFBConfigs, OpenGL
 version 2.1 instead of 1.5, but glxgears does not seem any faster
 (around 290 FPS). Other applications using GL (supertuxkart ...) still
 crash when trying to use hardware acceleration.
 
 For my part, I'd greatly appreciate *any* feedback from graphics people
 on this. Is it being worked on? Which component is the culprit (kernel,
 xorg driver, Mesa)?
 
 There is no DRM support for r600 and up in the kernel. Someone first has
 to finish his PhD first I believe...
 With the current kernel support you get _some_ benefits like xvideo, but
 that's about it for the moment.
 
 My 3850 (r600) has DRM support on Current:
 
 vga1 at pci1 dev 0 function 0 ATI Radeon HD 3850 rev 0x00
 wsdisplay0 at vga1 mux 1: console (80x25, vt100 emulation)
 wsdisplay0: screen 1-5 added (80x25, vt100 emulation)
 radeondrm0 at vga1: apic 2 int 16 (irq 10)
 drm0 at radeondrm0
 

No, not really. The bits that handle the instructions mesa sends to the
kernel are not in yet (*packet* functions iirc). Instructions for a r600
end up in the r300*packet* function and get rejected.  What works is
some blit stuff (2D) and Xvideo.

On some hw glxinfo will claim to have acceleration, but it will not
work. In that case, software rendering needs to be forced via an env
variable.

Re: How do I set process memory ulimits system wide on OpenBSD?

[Tobias Ulmer]

On Sat, Jan 01, 2011 at 02:54:48PM +, Douglas Held wrote:
 I've installed OpenBSD 4.7, i386 in a VMWare virtual machine with 3GB RAM.
 
 I find I can't allocate more than 1GB to any process as root.  ksh
 ulimit builtin provides me this when I try to set the hard limit
 unlimited.

1GB is the hard limit in the kernel (for i386). There are a number of
factors that play into this, the limitations of i386 with W^X, address
space randomisation, space for mmap, etc. Basically the price you pay
for OpenBSDs invisible security features.

There are some recent patches on tech@ that raise the limit a bit, iirc.

 
 Even so, when I set the hard and soft limits for, say, 'ulimit -d' as
 root and then su my application user, the specified limit is
 unattainable.
 
 # ulimit -d
 1048576
 # ulimit -Hd unlimited
 # ulimit -d unlimited
 # ulimit -d
 1048576
 # su - xyz
 $ ulimit -d
 524288
 $ ulimit -d 1024575
 ksh: ulimit: exceeds allowable limit
 
 Other operating systems have a configuration such as
 /etc/security/limits.conf.  What is the equivalent in OpenBSD?
 
 
 -- 
 Douglas Held
 d...@douglasheld.net
 +447986527654

Re: How do I set process memory ulimits system wide on OpenBSD?

[Tobias Ulmer]

On Sat, Jan 01, 2011 at 03:53:09PM +, Douglas Held wrote:
 OK. 1GB hard limit, I can work with that.
 
 What about the reduced limit for my non root user?  For now I'll
 simply carry out my processing as root, but this can hardly be
 considered best practices.

Put the user in the staff class (login.conf(5), passwd(5)). The user
can then raise its limits.

 
 Doug
 
 On Sat, Jan 1, 2011 at 3:23 PM, Tobias Ulmer tobi...@tmux.org wrote:
  On Sat, Jan 01, 2011 at 02:54:48PM +, Douglas Held wrote:
  I've installed OpenBSD 4.7, i386 in a VMWare virtual machine with 3GB RAM.
 
  I find I can't allocate more than 1GB to any process as root.  ksh
  ulimit builtin provides me this when I try to set the hard limit
  unlimited.
 
  1GB is the hard limit in the kernel (for i386). There are a number of
  factors that play into this, the limitations of i386 with W^X, address
  space randomisation, space for mmap, etc. Basically the price you pay
  for OpenBSDs invisible security features.
 
  There are some recent patches on tech@ that raise the limit a bit, iirc.
 
 
  Even so, when I set the hard and soft limits for, say, 'ulimit -d' as
  root and then su my application user, the specified limit is
  unattainable.
 
  # ulimit -d
  1048576
  # ulimit -Hd unlimited
  # ulimit -d unlimited
  # ulimit -d
  1048576
  # su - xyz
  $ ulimit -d
  524288
  $ ulimit -d 1024575
  ksh: ulimit: exceeds allowable limit
 
  Other operating systems have a configuration such as
  /etc/security/limits.conf.  What is the equivalent in OpenBSD?
 
 
  --
  Douglas Held
  d...@douglasheld.net
  +447986527654
 
 
 
 
 
 --
 Douglas Held
 d...@douglasheld.net
 +447986527654

Re: 4.8 arrival!

[Tobias Ulmer]

On Thu, Oct 28, 2010 at 09:58:22PM -0400, bsdmas...@hushmail.com wrote:
 Hello,
 
 Would you please consider uploading an iso image of your OpenBSD
 4.8 to some public tracker such as thepiratebay.org?
 
 If you are unfamiliar with the process of making an iso-image out
 of a CD, or if you need help with the generation and upload of the
 torrent file, I may be of some help. Just ask.
 
 Thanks alot, this will be of great use for poor folks like me who
 cannot afford the expensive license fees. Yes, I said it, 50CDN$ is
 very expensive. Maybe the OpenBSD Company could setup something
 like MSDNAA, for stuents to get access to the software for free?
 
 Anyway, I'm getting off topic.
 
 PS: please people, stop bottom-posting. It forces me to scroll down
 to read the latest message, and I don't like that. Show some common
 sense!
 
 
 On Thu, 28 Oct 2010 11:14:24 -0400 SJP Lists 
 sjp.li...@flashbsd.net wrote:
 On 27 October 2010 10:14, Rod Whitworth glis...@witworx.com 
 wrote:
  On Tue, 26 Oct 2010 17:36:00 -0500, Neal Hogan wrote:
 
 Chicago . . . THANKS!
 
 
  And all the way through customs to Sydney Australia.
  WOW!
 
 Me too.  And more nice shirts and a 2.5 CD for old times sake and 
 to
 get my hands on my favorite stickers!
 
 
 Shane



Obvious troll is obvious.

http://images.starcraftmazter.net/4chan/for_forums/failed_troll.jpg

bai

Re: libc glob issue?

[Tobias Ulmer]

On Fri, Oct 08, 2010 at 01:15:37AM -0400, Jeremy Chase wrote:
 I found this article that claims 4.7's ftpd and sftp are vulnerable to DoS:
 
 http://www.h-online.com/open/news/item/Flaw-in-libc-implementation-threatens-FTP-servers-1103319.html
 
 Which links to:
 
 http://ftp.netbsd.org/pub/NetBSD/security/advisories/NetBSD-SA2010-008.txt.asc
 
 
 I haven't seen any talk of it, is this an issue for us?

Yes

Re: OpenBSD Vim Programming FAQ

[Tobias Ulmer]

On Sat, Oct 09, 2010 at 01:27:21AM +0300, Vladimir Kirillov wrote:
[...]
 style(9):
 Indentation is an 8 character tab.  Second level indents are four spaces.
 
   while (cnt  20)
   z = a + really + long + statement + that + needs +
   two + lines + gets + indented + four + spaces +
   on + the + second + and + subsequent + lines;
 
 How do you guys follow this rule in vim neatly?
 

set cinoptions=:0,t0,+4,(4

Re: panic on 2 October and 3 October snapshots

[Tobias Ulmer]

On Sun, Oct 03, 2010 at 01:35:06PM -0500, eagir...@cox.net wrote:
 I installed the 2 October snapshot yesterday, and after the reboot got a 
 panic.

You installed? Not updated?

 
 I booted again from the RAMdisk kernel, point to the 3 October snaphot and 
 ignored the hash mismatch, and panicked again.  Trace and ps for each below.
 --
 Ed Ahlsen-Girard
 Ft. Walton Beach FL
 
 trace for 2 October snapshot::
 panic kernel diagnostic assertion flags  (M_WAITOK | M_NOWAIT) failed: \
 file ../../../../kern/kern_malloc.c line 194
 stopped at Debugger+0x4: popl %ebp
 Debugger(d08c4d72.dc17ac28,d08a370c,dc17ac28,0) at Debugger+0x4
 panic(d08a370c,d082b68e,d08ac09,d08a0fe4,c2) at panic+0x5d
 __assert(d082bb8e,d08a0fe4,c2,d08ac0c09,d6d5216a) at __assrt+0x2e
 malloc(1000,7f,0,0,11) at malloc+0x5e0
 end(1000,d09a93d0,30,1,1ff86) at 0xdc1928a6
 end(d1ec1180,1,2,d082a55e,ca81f) at 0xdc190e56
 lkmioctl(1c00,80044b03,dc17ae8c,3,d6d698a8,d6d52014,ca7e0,0) at \
 lkmioctl+0x48a

^ something related to loadable kernel modules

 spec_ioctl(dc17ada0,d6d52014,dc17adbc,d09a4b7f000) at \
 spec_ioctl+0x8c
 VOP_IOCTL(d6bdf174,80044b03,dc17ae8c,3,d667f000) at \
 VOP_IOCTL+0x42
 vn_ioctl(d6d5c948,80044b03,dc17ae8c,d6d698a8,4) at vn_ioctl+0x61
 sys_ioctl(d6d698a8,dc17af64,dc17af84,dc17afa8,d6d698a8) at \
 sys_ioctl+0x1b8
 syscall() at syscall+0x2f0
 syscall (number 0)
 0x2:
 
 ps for 2 October snapshot:
 PID PPIDPGRPUID S   FLAGS   WAITCOMMAND
 *19708  31790   31790   0   7   0x4000  modload

And there it is, you're loading a kernel module. You just voided your
warranty :)

 17681358135874  3   0x180   bpf pflogd
 13581   13580   3   0x80netio   pflogd
 14407   18189   18189   73  3   0x180   pollsyslogd
 18189   1   19198   0   3   0x88netio   syslogd
 31790   1   31790   0   3   0x4080  pause   sh
 13  0   0   0   3   100200  aiodonedaiodoned
 12  0   0   0   3   100200  syncer  update
 11  0   0   0   3   100200  cleaner cleaner
 10  0   0   0   3   100200  reaper  reaper
 9   0   0   0   3   100200  pgdaemonpagedaemon
 8   0   0   0   3   100200  bored   crypto
 7   0   0   0   3   100200  pftmpfpurge
 6   0   0   0   3   100200  usbtsk  usbtask
 5   0   0   0   3   100200  acpi0   acpi0
 4   0   0   0   3   100200  bored   syswq
 3   0   0   0   3   40100200idle0
 2   0   0   0   3   100200  kmalloc kmthread
 1   0   1   0   3   4080waitinit
 0   -1  0   0   3   80200   scheduler   swapper
 
 trace for 3 October snapshot::
 panic kernel diagnostic assertion flags  (M_WAITOK | M_NOWAIT) failed: \
 file ../../../../kern/kern_malloc.c line 194
 stopped at Debugger+0x4: popl %ebp
 Debugger(d08c4c1c.dc180c28,d08a35ac,dc180c28,60) at Debugger+0x4
 panic(d08a35ac,d082b56e,d08a0aa9,d08a0e84,c2) at panic+0x5d
 __assert(d082b56e,d08a0e84,c2,d08a0aa9,d6d6c8bc) at __assrt+0x2e
 malloc(1000,7f,0,0,11) at malloc+0x5e0
 end(1000,d09a93d0,30,1,1ff86) at 0xdc18e8a6
 end(d1e9d180,1,2,d082a43e,ca81f) at 0xdc18ae56
 lkmioctl(1c00,80044b03,dc180e8c,3,d6d698a8,d6d5f4b8,ca7e0,0) at \
 lkmioctl+0x48a
 spec_ioctl(dc180da0,d6d5f4b8,dc180dbc,d09a4b70,d6c20d04) at \
 spec_ioctl+0x8c
 VOP_IOCTL(d6c20d04,80044b03,dc180e8c,3,d6d7f000) at \
 VOP_IOCTL+0x42
 vn_ioctl(d6d5c9a0,80044b03,dc180e8c,d6d698a8,4) at vn_ioctl+0x61
 sys_ioctl(d6d698a8,dc180f66,dc180f84,d1da30e8,d6d79060) at \
 sys_ioctl+0x1b8
 syscall() at syscall+0x2f0
 syscall (number 0)
 0x2:
 
 
 PID PPIDPGRPUID S   FLAGS   WAITCOMMAND
 *13623  127012700   7   0x4000  modload
 24528   31806   31806   74  3   0x180   bpf pflogd
 31806   1   31806   0   3   0x80netio   pflogd
 12417   9937993773  3   0x180   pollsyslogd
 99371   99370   3   0x88netio   syslogd
 12701   12700   3   0x4080  pause   sh
 13  0   0   0   3   100200  aiodonedaiodoned
 12  0   0   0   3   100200  syncer  update
 11  0   0   0   3   100200  cleaner cleaner
 10  0   0   0   3   100200  reaper  reaper
 9   0   0   0   3   100200  pgdaemonpagedaemon
 8   0   0   0   3   100200  bored   crypto
 7   0   0   0   3   100200  pftmpfpurge
 6   0   0   0   3   100200  usbtsk  usbtask
 5   0   0   0   3   100200  acpi0   acpi0
 4   0   0   0   3   100200  bored   syswq
 3   0   0   0   3   40100200

Re: qemu core dumps

[Tobias Ulmer]

On Sun, Sep 12, 2010 at 04:28:25PM +0200, Elmar Bschorer wrote:
 hi list,
 
 I tried to install ubuntu with qemu as neither jconsole nor skype
 with emulation do work on openbsd 4.7 :-(

You are using the wrong operating system for the job. Don't.
There are plenty of Linux distros out there.

Re: Random core dumped with gtk+2 packages

[Tobias Ulmer]

On Wed, Aug 11, 2010 at 02:21:44AM +, Daniel B. wrote:
 Hi misc@,
 
 Recently, I'm having some problems running some packages, specially the 
 ones which use gtk+2, e.g. tagtool, easytag, and not so frequent, with 
 firefox too.

Correlation does not imply causation. easytag and the other tag tools
crash lots because the tag libraries they use are PoS. The GTK part of
Firefox is small and unimportant compared to the vast amount of buggy
Mozilla code.

 
 Get this after a few seconds:
 
 $ easytag
 Abort trap (core dumped)
 $
 
 I'm running -current. Tried to raise some resources with ulimit but with 
 no luck.
 
 Any clues? Thank you.

Run them in gdb and get a backtrace. Then hunt down the error and fix
it.

 
 My dmesg:
 OpenBSD 4.8 (GENERIC.MP) #0: Wed Aug 11 01:04:06 UTC 2010
 r...@sbc-fei02.my.domain:/usr/src/sys/arch/amd64/compile/GENERIC.MP
 real mem = 937295872 (893MB)
 avail mem = 898519040 (856MB)
 mainbus0 at root
 bios0 at mainbus0: SMBIOS rev. 2.5 @ 0xf (49 entries)
 bios0: vendor Phoenix Technologies, LTD version P09 date 05/05/2010
 bios0: FOXCONN M61PMV
 acpi0 at bios0: rev 0
 acpi0: sleep states S0 S3 S4 S5
 acpi0: tables DSDT FACP SSDT HPET MCFG SLIC APIC
 acpi0: wakeup devices HUB0(S5) XVR0(S5) XVR1(S5) XVR2(S5) UAR1(S5) USB0(S3) 
 USB2(S3) AZAD(S5) MMAC(S5)
 acpitimer0 at acpi0: 3579545 Hz, 24 bits
 acpihpet0 at acpi0: 2500 Hz
 acpimadt0 at acpi0 addr 0xfee0: PC-AT compat
 cpu0 at mainbus0: apid 0 (boot processor)
 cpu0: AMD Athlon(tm) II X2 240 Processor, 10408.49 MHz
 cpu0: 
 FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,MMX,FXSR,SSE,SSE2,HTT,SSE3,MWAIT,CX16,POPCNT,NXE,MMXX,FFXSR,LONG,3DNOW2,3DNOW
 cpu0: 64KB 64b/line 2-way I-cache, 64KB 64b/line 2-way D-cache, 1MB 64b/line 
 16-way L2 cache
 cpu0: ITLB 32 4KB entries fully associative, 16 4MB entries fully associative
 cpu0: DTLB 48 4KB entries fully associative, 48 4MB entries fully associative
 cpu0: apic clock running at 200MHz
 cpu1 at mainbus0: apid 1 (application processor)
 cpu1: AMD Athlon(tm) II X2 240 Processor, 2812.98 MHz
 cpu1: 
 FPU,VME,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,MMX,FXSR,SSE,SSE2,HTT,SSE3,MWAIT,CX16,POPCNT,NXE,MMXX,FFXSR,LONG,3DNOW2,3DNOW
 cpu1: 64KB 64b/line 2-way I-cache, 64KB 64b/line 2-way D-cache, 1MB 64b/line 
 16-way L2 cache
 cpu1: ITLB 32 4KB entries fully associative, 16 4MB entries fully associative
 cpu1: DTLB 48 4KB entries fully associative, 48 4MB entries fully associative
 ioapic0 at mainbus0: apid 4 pa 0xfec0, version 11, 24 pins
 ioapic0: misconfigured as apic 0, remapped to apid 4
 acpiprt0 at acpi0: bus 0 (PCI0)
 acpiprt1 at acpi0: bus 1 (HUB0)
 acpicpu0 at acpi0: PSS
 acpicpu1 at acpi0: PSS
 acpitz0 at acpi0: critical temperature 70 degC
 acpibtn0 at acpi0: PWRB
 pci0 at mainbus0 bus 0
 NVIDIA MCP61 Memory rev 0xa1 at pci0 dev 0 function 0 not configured
 pcib0 at pci0 dev 1 function 0 NVIDIA MCP61 ISA rev 0xa2
 nviic0 at pci0 dev 1 function 1 NVIDIA MCP61 SMBus rev 0xa2
 iic0 at nviic0
 spdmem0 at iic0 addr 0x50: 1GB DDR2 SDRAM non-parity PC2-6400CL5
 iic1 at nviic0
 iic1: addr 0x2f 01=33 02=33 03=33 04=33 05=33 06=33 07=33 08=33 09=60 0a=3f 
 0b=57 0c=98 11=70 12=40 5a=05 5b=10 5c=10 5d=19 5e=34 f0=20 f1=20 f2=02 fa=00 
 ff=5e words 00= 01=33ff 02=33ff 03=33ff 04=33ff 05=33ff 06=33ff 07=33ff
 NVIDIA MCP61 Memory rev 0xa2 at pci0 dev 1 function 2 not configured
 ohci0 at pci0 dev 2 function 0 NVIDIA MCP61 USB rev 0xa3: apic 4 int 5 (irq 
 5), version 1.0, legacy support
 ehci0 at pci0 dev 2 function 1 NVIDIA MPC61 USB rev 0xa3: apic 4 int 10 
 (irq 10)
 usb0 at ehci0: USB revision 2.0
 uhub0 at usb0 NVIDIA EHCI root hub rev 2.00/1.00 addr 1
 ppb0 at pci0 dev 4 function 0 NVIDIA MCP61 rev 0xa1
 pci1 at ppb0 bus 1
 vr0 at pci1 dev 7 function 0 VIA VT6105 RhineIII rev 0x86: apic 4 int 5 
 (irq 5), address 00:08:54:19:6a:a2
 ukphy0 at vr0 phy 1: Generic IEEE 802.3u media interface, rev. 4: OUI 
 0x004063, model 0x0034
 rl0 at pci1 dev 8 function 0 Realtek 8139 rev 0x10: apic 4 int 11 (irq 11), 
 address 00:e0:7d:a9:13:06
 rlphy0 at rl0 phy 0: RTL internal PHY
 azalia0 at pci0 dev 5 function 0 NVIDIA MCP61 HD Audio rev 0xa2: apic 4 int 
 5 (irq 5)
 azalia0: codecs: VIA/0xe721
 audio0 at azalia0
 pciide0 at pci0 dev 6 function 0 NVIDIA MCP61 IDE rev 0xa2: DMA, channel 0 
 configured to compatibility, channel 1 configured to compatibility
 pciide0: channel 0 disabled (no drives)
 pciide0: channel 1 ignored (disabled)
 nfe0 at pci0 dev 7 function 0 NVIDIA MCP61 LAN rev 0xa2: apic 4 int 15 (irq 
 15), address 00:23:ae:ff:de:0b
 brgphy0 at nfe0 phy 0: BCM54XX 10/100/1000baseT PHY, rev. 1
 pciide1 at pci0 dev 8 function 0 NVIDIA MCP61 SATA rev 0xa2: DMA
 pciide1: using apic 4 int 11 (irq 11) for native-PCI interrupt
 wd0 at pciide1 channel 0 drive 0: SAMSUNG HD161GJ
 wd0: 16-sector PIO, LBA48, 152627MB, 312581808 sectors
 wd0(pciide1:0:0): using PIO mode 4, Ultra-DMA mode 6
 atapiscsi0 at pciide1 channel 

Re: DRM/OpenGL problems with Radeon HD 4670 on -current

[Tobias Ulmer]

On Mon, Aug 09, 2010 at 08:06:31PM +0200, Mattieu Baptiste wrote:
 Hi all,
 
 I have some problems I am seeing with DRM on my Radeon HD
 4670 in -current (dual head setup with two monitors at
 1280x1024). I can't display any OpenGL applications.
 
 The best way to reproduce this is by running glxgears :
 
 matt...@kronenbourg: ~ $ glxgears
 drmRadeonCmdBuffer: -22. Kernel failed to parse or rejected command
 stream. See dmesg for more info.
 
 
 In dmesg I have these errors :
 
 error: [drm:pid30830:r300_emit_carefully_checked_packet0] *ERROR*
 Register 4e4c failed check as flag=00
 error: [drm:pid30830:r300_do_cp_cmdbuf] *ERROR* r300_emit_packet0 failed

Expected behaviour. export LIBGL_ALWAYS_SOFTWARE=1 to revert to software
rendering

Re: hardware ports suggestion

[Tobias Ulmer]

On Sat, Jul 24, 2010 at 02:10:57AM -0700, Philip Guenther wrote:
 2010/7/24 sERGEJ bRONNIKOW este...@gmail.com:
  My idea is to make ports for such equipment.
 
 Lots of people have lots of really good ideas that don't get done
 because no one is willing to spend time or money on them.  Are you
 willing to spend *your* time and *your* money on your idea?  Because
 if not, all of the OpenBSD developers have *plenty* of their own good
 ideas to work on.
 
 If you're serious and not just talking, then have you read all the
 pages in the OpenBSD website that describe how the various ports that
 *do* exist came about...and about how those that are incomplete,
 unmaintained, or dead came to be that way?  Do you have a reason (and
 not just an idealogy!) for why *this* port (to the unspecified
 platform you have in your mind) will succeed where others have been
 left behind?
 
 
 If so, well, please send a patch for /usr/src/sys/arch/whatever,
 plus three machines that can run it (one for builds, one for ports,
 and one for some other developer to work with) to the project.  If you
 don't have code or machines than you're just talk.  Come back when you
 have one, the other, or both.

I've read his mail twice and I think what he wants is to create ports
(as in ports tree) with hardware schematics/instructions/sources and
maybe firmware files for small DIY uC projects.

I would welcome such a thing, if the project is possible and achievable
with OpenBSD. CAD/cross-compilers/etc would have to be available and
usb/serial/jtag communication must work (no point of getting such a port
into OpenBSD if only Linux has the specialised usb driver for flashing
the hw)

Of course talk is cheap, he should make a port we can look at first...

Tobias

 
 
 Philip Guenther

Re: Any ideas on this crash?

[Tobias Ulmer]

On Mon, Jun 21, 2010 at 03:41:21PM -0400, STeve Andre' wrote:
My package builder died this weekend when I couldn't get to it.
 I may have hardware problems--I'm not sure.  Below is the relevent
 data typed in.  Any ideas?  This is an i386-current system compiled
 on June 15th.
 
 Thanks, STeve Andre'
 
 - ps  trace data 
 First bad
 /: bad dir ino 14 at offset 69632: mangled entry
 panic: bad dir

Says it right there, disk structure corrupted. A fsck run will fix it.

Re: USB Controller Causing Issues

[Tobias Ulmer]

On Fri, May 07, 2010 at 10:11:56AM -0700, Ben Niccum wrote:
 I have a couple issues using OpenBSD with a VIA VB7001G motherboard
 focused around USB devices. This board has been tested on both 4.6 and
 4.7-current (April 22nd Snapshot).
 
 First of all, whenever I attempt to use a bootable OpenBSD USB flash
 disk (created using the preferred method described here:
 http://openbsd.org/faq/faq14.html#flashmemLive), the motherboard fails
 to POST. It will begin an attempt and then continually try to reboot.
 
 This same USB flash disk has been tested on other motherboards and
 works without any issues. I have also tried testing with three different
 USB storage devices.
 
 The second issue is that whenever there is a USB storage device plugged
 into the motherboard the boot process freezes. I have included a dmesg
 below from a normal boot and indicated the point at which booting
 freezes.
 
 If I wait until after the system has booted and then plug in a USB
 storage device, it freezes at the moment the device was plugged in.
 
 This has been tested on devices with filesystems that are OpenBSD (A6),
 FreeBSD (A5) and MSDOS (0B). Again this has been tested on three
 different devices.
 
 To emphasize, this only occurs for USB *storage devices*. USB mice,
 keyboards, ethernet adapters, cdrom-drives, camera etc. do not cause
 this issue.
 
 I have two identical VIA VB7001G motherboards and both of them produce
 the exact same results in all the cases I have described above.
 
 Here is the dmesg from the system, obtained by having no USB storage
 devices attached. The last line printed by OpenBSD prior to all
 instances of the OS freezing is marked by '*'s.
 
 --- Begin DMESG ---
 OpenBSD 4.7-current (GENERIC) #620: Thu Apr 22 11:50:48 MDT 2010
 dera...@i386.openbsd.org:/usr/src/sys/arch/i386/compile/GENERIC
 cpu0: VIA Esther processor 1500MHz (CentaurHauls 686-class) 1.50 GHz
 cpu0:
 FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,CMOV,PAT,CFLUSH,ACPI,MMX,FXSR,SSE,SSE2,TM,SBF,SSE3,TM2
 real mem  = 1005023232 (958MB) avail mem = 963829760 (919MB)
 mainbus0 at root
 bios0 at mainbus0: AT/286+ BIOS, date 11/19/07, BIOS32 rev. 0 @
 0xf9440, SMBIOS rev. 2.3 @ 0xf (24 entries) apm0 at bios0: Power
 Management spec V1.2 (slowidle) apm0: AC on, battery charge unknown
 acpi at bios0 function 0x0 not configured
 pcibios0 at bios0: rev 2.1 @ 0xf/0xc594
 pcibios0: PCI IRQ Routing Table rev 1.0 @ 0xfc500/144 (7 entries)
 pcibios0: bad IRQ table checksum
 pcibios0: PCI BIOS has 7 Interrupt Routing table entries
 pcibios0: PCI Exclusive IRQs: 5 10 11
 pcibios0: PCI Interrupt Router at 000:17:0 (VIA VT8237 ISA rev 0x00)
 pcibios0: PCI bus #1 is the last bus
 bios0: ROM list: 0xc/0xfc00 0xd/0x2000

Try setting Plug and play OS in the BIOS to yes.
Disabling apm via ukc could also help.

Re: thinkpad windows refund

[Tobias Ulmer]

On Sun, Apr 18, 2010 at 02:59:25PM -0400, RALOVICH, Kristsf wrote:
 Dear All,
 
 I would like to ask Thinkpad or Lenovo machine owners on the mailing
 list if they had any experience on returning and receiving a refund
 for windows bundled with newly bought machines in the US or Canada.

http://translate.google.com/translate?hl=ensl=datl=enu=http%3A%2F%2Fphk.freebsd.dk%2FMicrosoftSkat%2F

Try it, bug prepare to sue...

 
 Thanks,
 Kristof

Re: which ISO for a VM?

[Tobias Ulmer]

On Fri, Apr 02, 2010 at 05:23:25AM -0400, Zachary Uram wrote:
 I have never run OpenBSD before and want to try it out. Wondering if
 there is an ISO I can run in VirtualBox? If not what is the

VB is broken. They're patching the kernel on the fly and seem to miss
the right spot depending on the phase of the moon or whatever. Do not
blame any issues on OpenBSD.

 recommended method for users who wish to run OpenBSD in
 virtualization?

You may try vmware. cdXX.iso is enough.

 
 Regards,
 Zach
 
  http://www.fidei.org 

Re: Problem after upgrade 4.5 to 4.6: ERR M

[Tobias Ulmer]

On Mon, Mar 22, 2010 at 04:41:39PM +0800, Uwe Dippel wrote:
 Having done upgrades from 4.0 onwards, on a OpenBSD-only server
 (amd64), this time something must have gone wrong: Despite of the
 (remote, I have no physical access, via serial console) 'successful'
 upgrade (no error messages), when I was asked to reboot, I did, as
 always. Alas, it came up with
 
 Attempting Boot From Floppy Drive (A:)
 Attempting Boot From CD-ROM
 Attempting Boot From Hard Drive (C:)
 Using drive 0, partition 3.
 Loading...
 ERR M

That is biosboot(8) telling you that it cannot find /boot, which is the
boot loader that prints the boot prompt and so on. Biosboot has to be
really small, so its error message are terse.

 
 on an HP ML350G4p.
 
 From all I know it is a problem with the MBR.

As explained above, no, you likely moved around/corrupted /boot in a way
that doesn't work for biosboot.

 What I'd really like to get, before I drive there and get access, is
 how to best solve this problem, and most straightforward. Talking
 about what went wrong can wait, since this is a production machine
 and should be back as soon as possible.

The procedure is straight forward and documented in installboot(8).

 
 Thanks in advance,
 
 Uwe

Re: Problem after upgrade 4.5 to 4.6: ERR M

[Tobias Ulmer]

On Mon, Mar 22, 2010 at 09:59:20PM +0800, Uwe Dippel wrote:
[..]
 
 Thanks for the reply. I'll go there next to try what has been
 proposed. Before I try, in case the
 # /usr/*m*dec/installboot -v boot /*usr/mdec*/biosboot sd0
 does NOT work, what else could I do? (I am asking, because it is a
 server room quite far away, with little chance for me to
 communicate, and difficult to go.) So, is there any alternative, or
 additional, solution to fall back to, when I am there, and
 installboot doesn't cut it?
 
 Uwe

Well, re-install. The installer does it the same way.

Re: kde4 dead?

[Tobias Ulmer]

On Mon, Mar 15, 2010 at 11:00:54PM +0100, Marc Espie wrote:
 On Mon, Mar 15, 2010 at 03:19:41PM +0100, Tobias Ulmer wrote:
  On Mon, Mar 15, 2010 at 04:33:03AM -0500, Donald Cooley wrote:
   openports shows that the openbsd version of kde4 is nearly two years
   old.  are there any future plans to update kde4?
   
   Regards,
   Donald Cooley
  
  http://lists.kde.org/?l=kde-develw=2r=1s=openbsdq=b
  KDE doesn't give a fuck about OpenBSD, so why should we?
 
 You're totally mistaken.
 
 KDE cares about the BSDs, and they're very much no-nonsense people.
 I had absolutely no difficulty getting an account with them, nor with
 folding back portable patches I had to make things work on OpenBSD.

KDE cares in so far as they accept patches and would welcome a developer
that targets OpenBSD. I know because I've talked to some of them on IRC.

There's however no effort to do it themselves or even set up a testbox
to make sure all their developers would notice that their applications
don't work on !Linux.

 
 The main reason we're behind for kde4 is that it's mostly impossible
 to compile kde4 with gcc3, so there is some upheaval there.

I'm aware of that. To be honest, I thought a little oil in the fire
can't hurt and could even attract some responses that may change the
situation ;-)

 
 Also the fact that back when I ported kde 4.0, it was not interesting
 at all, especially compared to 3.5.10.
 
 Other issues have happened since then. It shouldn't be that hard to get
 kde4 to work, once you get past the gcc4 issue (and port cmake, but
 apparently 2.8.0 is nicer).

Re: kde4 dead?

[Tobias Ulmer]

On Mon, Mar 15, 2010 at 04:33:03AM -0500, Donald Cooley wrote:
 openports shows that the openbsd version of kde4 is nearly two years
 old.  are there any future plans to update kde4?
 
 Regards,
 Donald Cooley

http://lists.kde.org/?l=kde-develw=2r=1s=openbsdq=b
KDE doesn't give a fuck about OpenBSD, so why should we?

Re: Problems with Build World

[Tobias Ulmer]

On Tue, Mar 02, 2010 at 03:15:30AM -0600, Ron McDowell wrote:
 My script builds the 4.6-release and 4.6-stable pulled 1/26.  It
 chokes on 2 copies of 4.6-stable pulled today.
 
 As I've said, I don't care about the error at this point, I want to
 know how the build process works.  Restating my question:

build:
.ifdef GLOBAL_AUTOCONF_CACHE
cp /dev/null ${GLOBAL_AUTOCONF_CACHE}
.endif
cd ${.CURDIR}/share/mk  exec ${SUDO} ${MAKE} install
cd ${.CURDIR}/include  ${MAKE} prereq  exec ${SUDO} ${MAKE} includes
${SUDO} ${MAKE} cleandir
cd ${.CURDIR}/lib  ${MAKE} depend  ${MAKE}  \
NOMAN=1 exec ${SUDO} ${MAKE} install
cd ${.CURDIR}/gnu/lib  ${MAKE} depend  ${MAKE}  \
NOMAN=1 exec ${SUDO} ${MAKE} install
${MAKE} depend  ${MAKE}  exec ${SUDO} ${MAKE} install

Re: softdeps enabled = poor concurrent access?

[Tobias Ulmer]

On Wed, Feb 24, 2010 at 10:52:05AM +0100, Jan Stary wrote:
 On Feb 23 19:20:28, Noah McNallie wrote:
  Hey guys. Noah here. I'd like to use openbsd on an older machine i have.  
  I've had it on there before and never tested something that i've been  
  testing on various operating systems lately. That's how well they do  
  while under disk io load, concurrently.
 
  An example would be to tar -zxvf a large tarball and in another  
  terminal, try to run a simple command. such as 'uname' or 'ls' or what  
  have you. To test responsiveness. It may not be a very good test but  
  it's a everyday usage test.
 
  Well, i've found on openbsd without sofdeps enabled it will do this just  
  fine. But when enabling softdeps it will not. The 'uname' or 'ls' will  
  take quite a while to complete.
 
 So, your system is slow _with_ softupdates?

No, softdep trades responsiveness for speed. The overall throughput increases,
of course. I've discussed this with him on IRC some time ago. It was futile,
he's hung up on thinking he can get the impossible by whining about it.

This thread is as pointless as anything.

 
  The machine is a 300MHz 2MB L2 sparc64 SUN Ultra 30. softdeps is almost  
  required as it speeds up something like the extraction of a tarball  
  exponentially. I'm guessing somewhere near 25x. It's very slow on this  
  machine without sofdeps.
 
 So, your system is slow _without_ softupdates?

Re: OT, .. but has anyone seen a crontab editor

[Tobias Ulmer]

On Fri, Feb 19, 2010 at 04:14:44PM -0600, L. V. Lammert wrote:
 On Sat, 20 Feb 2010, Paul M wrote:
 
  it's **Not clear whatproblem you're actualy trying to solve.**
 
 What's so difficult about need a way to edit crontab with something like
 an nCurses interface? That seems to be, by definition, simple,
 point-and-click, definate options, no man pages, no vi editors, ...

vi(1) is a ncurses interface.

 
  I can imagine a situation where your question is valid and sensible, but
  that would be just be me going off on a tangent - give us some
  background, explain *properly* why the answers you've been given are
  unsatisfactory.
  I have to say, it does sound to me as if you're being deliberately
  obtuse.
 
 Certainly not intended, .. however I cannot imagine why the statment above
 does not describe the problem accurately  succinctly.
 
  Give us the full story, and I'm sure you'll get a very good answer. Or
  several.
 
 The chaps tweaking the crontab entries are Windoze admins, and they need
 to adjust the start/stop times on cronjobs that start and stop replication
 services. It would *seem* that there would be a way to apply all this
 fancy technology we have in our toolkits for a simple, point-and-shoot (a
 la nCurses) UI that requires no a priori knowledege other than an account
 name  password.
 
   Lee

In the time you've been spamming my inbox, every half-competent sysadmin
could have learned ncurses(3) and write the perfect(tm) interface for
his purpose.

I'll just leave this here:
http://doxfer.com/Webmin/ScheduledCommands#The_Scheduled_Commands_module

Re: Latest snapshot doesn't work in Qemu under Fedora 12

[Tobias Ulmer]

Uhm, looks seriously off-topic on any mailinglist that ends in
@openbsd.org. I hope you reported this to Fedora.

On Fri, Feb 12, 2010 at 01:25:11PM +0100, Tomas Bodzar wrote:
 [ backtrace of crash in hand-rolled Drepper assembly ]

Re: ATI Device Documentation - Evergreen

[Tobias Ulmer]

On Mon, Feb 01, 2010 at 08:19:42PM -0600, Axton wrote:
 If these docs are in line with what is needed to develop a usable driver and
 there are any developers @openbsd.org out there interested in developing a
 driver for this card and in need of a hardware donation, let me know.
 
 http://developer.amd.com/gpu/ATIStreamSDK/assets/AMD_Evergreen-Family_ISA_Instructions_and_Microcode.pdf
 
 - Axton Grams

That's an ISA manual, basically describes which instructions the GPU
understands and what effect they have. This is very useful for people
who want to write accelerated drivers etc, but not for integration in
an operating system. There's probably another manual that describes the
registers, initialisation, memory etc.

GPU driver development is usually done by Xorg. The Xorg mailinglist is
probably a better place to ask. If you do, include more information
about your actual hardware.

Btw. all the relevant developers are aware that ATI publishes these
documents. The real issue is the lack of time or money. If you can help
them on that front, that would be great...

Re: Postgresql and Memory Usage

[Tobias Ulmer]

On Thu, Jan 28, 2010 at 06:53:06PM -0500, Ted Unangst wrote:
 On Thu, Jan 28, 2010 at 6:10 PM, Tobias Ulmer tobi...@tmux.org wrote:
  Let's make this simple, on i386 you have 1G per process. Adding all the
  numbers up, you have to stay below this limit.
 
 That's quite an oversimplification.  For starters, we're talking about
 shm, which doesn't count against your 1GB.  The stack doesn't count.
 Libraries don't count.  File backed mmap doesn't count.  In fact, the
 set of numbers that must add up to less than 1GB only contains one
 number.

OK ok, maybe it's time for an undeadly article? ;)

Re: Postgresql and Memory Usage

[Tobias Ulmer]

On Wed, Jan 27, 2010 at 02:13:45PM -0700, Jeff Ross wrote:
 I have searched (and searched) so I wonder if I'm running into the
 i386 1GB limit I see referenced, as in the thread today about fsck
 on larger partitions.

Yes you do. Also, kernel memory is limited, insane shm value will
probably (havn't looked at the code) have bad effects.

Re: fsck segfault on a big partition, 4.6

[Tobias Ulmer]

On Wed, Jan 27, 2010 at 12:38:47AM +, Rob Sheldon wrote:
 Hi,
 
 So, the short version is that I have a server with OpenBSD 4.6 that can't
 fsck its big partition; fsck fails with a segfault every time. If I ulimit
 -d unlimited before fsck'ing, it just takes a little longer to segfault.
 It produces no other output. IIRC, the partition is roughly 6 TB. Two
 questions then: is there any way through this that doesn't involve
 newfs'ing the partition, and is there a right way to do a partition of
 that size in OpenBSD given fsck's 1G hard limit?

Amd64 allows 8G. Increase newfs blocksize to 64k (make sure you don't
run out of inodes), that should lessen the memory requirements a bit
and make fsck runs a little faster.

I have my doubts about OpenBSD as a (backup) file server with large
filesystems, there might be a more appropriate OS for the job.

Re: Books on reverse engineering?

[Tobias Ulmer]

On Wed, Jan 20, 2010 at 05:52:52PM -0800, James Hozier wrote:
 With every single laptop I've bought/been given over the years, I
 was able to run OpenBSD on them almost flawlessly save a few
 quick/simple hacks to make anything that didn't work, work.
 
 The one main issue I've had with ALL of them was the wireless
 card...maybe I was just unlucky to have gotten ones with crappy
 chipsets (like this Broadcom I have now which is totally useless...
 I want to stomp on it real badly) but nonetheless it pisses me off.
 
 I want to try and help solve my own problems as well as for the OBSD
 community who might also have this particular issue, so I'm looking
 to research on how to reverse engineer these things and write drivers
 for them.
 
 I know it's not easy, even though I don't understand how hard it is
 because I've never done it before, but I do hear that if there's a
 hell, it's a place where people are sent to do this for eternity.
 
 So with that reference in mind, would anyone experienced care to point
 me in some correct direction? (Which texts to read, which programming
 language(s) to focus on, etc.)

- C
- any intro/boot to x86 assembly; to get the basics
- intel cpu pdfs
- ida pro / ollydbg
- something on computer architecture.
- windows ddk to get an idea how drivers work on windows, possibly book
  on same topic.
- BSD basics (McKusick, Bach, etc) + whatever you can get your hands on
- Device is connected via a BUS to CPU - docs.
- IEEE standards
- any other docs.
- more of the same
- Read lots of code.
- supertanker sized amounts of experience
- ability to research stuff yourself, without asking on a ml
- etc

Your question is naive. If you were up to it, you wouldn't have to ask
the equivalent of How do I become an awesome hacker?.

Writing this up was and is a waste of time, it will never happen.

Re: Books on reverse engineering?

[Tobias Ulmer]

The only one who can prove that my assumptions are BS would be James.
The pressure is on, maybe you want to help him with better pointers than
mine instead of just calling bs.

On Thu, Jan 21, 2010 at 10:19:10PM +, Owain Ainsworth wrote:
 As someone who went from knowing a small amount of C  to hacking the
 kernel, i call bullshit on your assumptions here.
 
 On 1/21/10, Tobias Ulmer tobi...@tmux.org wrote:
  On Wed, Jan 20, 2010 at 05:52:52PM -0800, James Hozier wrote:
  With every single laptop I've bought/been given over the years, I
  was able to run OpenBSD on them almost flawlessly save a few
  quick/simple hacks to make anything that didn't work, work.
 
  The one main issue I've had with ALL of them was the wireless
  card...maybe I was just unlucky to have gotten ones with crappy
  chipsets (like this Broadcom I have now which is totally useless...
  I want to stomp on it real badly) but nonetheless it pisses me off.
 
  I want to try and help solve my own problems as well as for the OBSD
  community who might also have this particular issue, so I'm looking
  to research on how to reverse engineer these things and write drivers
  for them.
 
  I know it's not easy, even though I don't understand how hard it is
  because I've never done it before, but I do hear that if there's a
  hell, it's a place where people are sent to do this for eternity.
 
  So with that reference in mind, would anyone experienced care to point
  me in some correct direction? (Which texts to read, which programming
  language(s) to focus on, etc.)
 
  - C
  - any intro/boot to x86 assembly; to get the basics
  - intel cpu pdfs
  - ida pro / ollydbg
  - something on computer architecture.
  - windows ddk to get an idea how drivers work on windows, possibly book
on same topic.
  - BSD basics (McKusick, Bach, etc) + whatever you can get your hands on
  - Device is connected via a BUS to CPU - docs.
  - IEEE standards
  - any other docs.
  - more of the same
  - Read lots of code.
  - supertanker sized amounts of experience
  - ability to research stuff yourself, without asking on a ml
  - etc
 
  Your question is naive. If you were up to it, you wouldn't have to ask
  the equivalent of How do I become an awesome hacker?.
 
  Writing this up was and is a waste of time, it will never happen.
 
 
 
 -- 
 Sent from my mobile device

Re: writing to usb very slow

[Tobias Ulmer]

On Wed, Jan 20, 2010 at 12:34:19AM +0100, T. Tofus von Blisstein wrote:
 # mount -o async /dev/sd1a /mnt
 # date  cp -r TEST/ /mnt  umount /mnt  date
 Wed Jan 20 00:17:54 CET 2010
 Wed Jan 20 00:33:35 CET 2010

And you think because of all this noise, someone is going to sit down
and fix it for you? Boy are you wrong...

 
 
 2010/1/20 David Vasek va...@fido.cz:
  On Tue, 19 Jan 2010, T. Tofus von Blisstein wrote:
 
  Hello David,
 
  thanks.
 
  You're welcome.
 
  # mount /dev/sd1a /mnt
  # date  cp -r TEST/ /mnt  umount /mnt  date
  Tue Jan 19 23:11:27 CET 2010
  Tue Jan 19 23:29:12 CET 2010
 
  So it's reduced a lot, but still it is much slower than... sorry guys,
  the penguin.
 
  Still, from 50 minutes to 18 to copy 256M is a significant improvement
 
  Try using softdep option for mount, possibly even the async option, if
  you like the risk, otherwise you are comparing two completely different
  modes of filesystem operation.
 
  Regards,
  David

Re: obsd as dom0?

[Tobias Ulmer]

On Sun, Jan 10, 2010 at 12:28:02PM +0100, Vadkan Jozsef wrote:
 Is it possible?

No.

Re: Xorg -br option does not work anymore

[Tobias Ulmer]

On Sat, Dec 19, 2009 at 10:56:59AM -0500, Philippe Meunier wrote:
 Hello,
 
 Xorg's -br option does not seem to work anymore.  When I try it I get
 the standard X grey pattern on the root window instead of getting
 solid black.  The option '-nolisten tcp' still works, and I have not
 tried to test other options.  I noticed the change after upgrading a
 desktop PC from 4.5-current to 4.6-current about a month and a half
 ago (recompiling OpenBSD and Xenocara from source) and saw the same
 change again yesterday when doing the same upgrade using the same
 homemade release on a Thinkpad laptop (T43).  Can anyone confirm this
 or is it just me?  I know about 'xsetroot -solid black', I just would
 like to know whether this is an Xorg bug or a problem with the way my
 machines are configured and / or upgraded.

-br is the default now. Some developers don't seem to be pleased with that
change, which is why OpenBSD patches it. This breaks -br.

http://cgit.freedesktop.org/xorg/xserver/commit/?id=0bb317a78b96fddcdac319c9706b3a12f931ea44
http://www.openbsd.org/cgi-bin/cvsweb/xenocara/xserver/os/utils.c.diff?r1=1.8;r2=1.9;f=h

Sending a diff is your job now :)

 
 $ Xorg -version
 [...]
 X.Org X Server 1.6.3.901 (1.6.4 RC 1)
 Release Date: 2009-8-25
 X Protocol Version 11, Revision 0
 Build Operating System: OpenBSD 4.6 i386 
 Current Operating System: OpenBSD akpatok.ungava.bay 4.6 GENERIC#4 i386
 Build Date: 28 October 2009  04:45:26PM
 [...]
 $ dmesg | head -1
 OpenBSD 4.6-current (GENERIC) #4: Wed Oct 28 15:35:02 ICT 2009
 
 Thanks,
 
 Philippe

Re: Inside Out Networks Edgeport USB Serial Adapters

[Tobias Ulmer]

On Wed, Dec 16, 2009 at 04:58:12PM -0600, Brad DeMorrow wrote:
 Hey guys, I need a bit of help if you don't mind.
 
 I've been trying to hack up a basic driver to communicate with my Edgeport
 device..
 It started out decent enough, I got a new driver I named uep(for no
 particular reason besides it looked available) to attach to my device just
 fine..
 So at this time I have an extremely simple driver with just these
 functions..
 -uep_match
 -uep_attach
 -uep_detach
 
 ..
 
 Anyway, to get to my point, I read about lkm and thought man that sure would
 be nice if I didn't have to reboot my machine every time I make a change to
 my sources...
 
 Anyway, I was getting
 modload: entry point _uep_lkmentry not found in uep.o
 
 every time I tried to load my module, and I thought for sure I was doing
 something stupid, but it turns out that even the examples inside
 /usr/share/lkm give the same error when I try to load them.

Nice todo for the people who always complain that there is no list of
things to fix in OpenBSD.

 
 I can't find very much documentation on this, so I was hoping someone could
 point me in the right direction..

These examples worked 15 months ago, and they still should...
http://tin.tmux.org/~tobiasu/hgweb/lkm/file/

 
 FYI - I am of course running -current kernel and userland(as of maybe 2 days
 ago).
 
 Thanks,

Re: DRI not loading on ATI Radeon Mobility 7500/M7 (4.6-RELEASE)

[Tobias Ulmer]

On Thu, Dec 10, 2009 at 03:36:09AM -0500, Anders Langworthy wrote:
 According to radeon(4), my ATI Radeon Mobility 7500/M7 is supported by
 the driver but as far as I can tell DRI is not loading properly when I
 start X.  X seems to work well enough besides 3D performance, which is
 dreadful.  An incriminating bit seems to be here:
 
 ...
 (II) RADEON(0): initializing int10
 (II) RADEON(0): Primary V_BIOS segment is: 0xc000
 (II) RADEON(0): Legacy BIOS detected
 drmOpenDevice: node name is /dev/drm0
 drmOpenDevice: open result is -1, (Operation not permitted)

Are you in the wheel group? See permissions of /dev/drm0

Re: error when updating ports

[Tobias Ulmer]

On Tue, Dec 08, 2009 at 07:43:00AM +0200, Ismail OZATAY wrote:
 checking for gcc... /usr/ports/obj/gcc-4.2.4/bin/egcc
 checking whether the C compiler (/usr/ports/obj/gcc-4.2.4/bin/egcc
 -O2 -g ) works... no
 configure: error: installation or configuration problem: C compiler
 cannot create executables.

You break it, you fix it.

Re: SATA CDRW/DVDRW

[Tobias Ulmer]

On Sun, Dec 06, 2009 at 11:55:59AM -0600, Chris Bennett wrote:
 Do OpenBSD 4.6 and/or current support SATA for CD/DVD?

Yes it does, mine works just fine. But the area seem to be pretty hairy

 I checked and chipset is supported
 I also turned of RAID mode, didn't stop crashes (which could be
 something else anyway)

sendbug(1)
 
 Thanks