Re: httpd setup info?

>>What's the appropriate way to let the browser >> know it should open it in Acrobat See "Content-Disposition" header. https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Content-Disposition It tells client to download document or open it inline.

Re: FAQ: dmesg archive

See Otto's comment https://marc.info/?l=openbsd-misc=117169914814561=2 On Tue, Jun 26, 2018 at 3:59 PM, Rupert Gallagher wrote: > There seems to be a dm...@openbsd.org address where to post such stuff, > but I could not find its archive, nor I could find a searchable database. > Do I have to

Re: how to know the progressive state of dd

I do not understand what are you trying to achieve, but instead of killall you may use pkill(1) On Tue, Jun 26, 2018 at 1:33 AM, Tuyosi T wrote: > hi all . > > on Linux > > dd-progress.bat < > --- > while true > do > date > killall -USR1

Re: Partitioning recommendations for 6.3?

Do you want to really build all ports or just fetch skeletons and build some of them? For skeletons, automatic layout is good enough, but I recommend to increase /usr/src a little and decrease /home. Make sure you have ~ 5GB for /usr/src/ and /usr/obj. On Mon, Jun 25, 2018 at 3:17 PM, John

Re: Have a 1 Gbit connection but I not all devices are getting 1 Gbit speed from my ISP

> > > >Lenovo G500 with USB 3.0 Ethernet and OpenBSD 6.2 - 100mbit > >Lenovo T400 with motherboard Ethernet Ubuntu 14.04 - 100mbit > >Lenovo T560 with motherboard Ethernet Windows 10 - 1 Gbit > > > Boot all 3 laptops with OpenBSD flashdrive to see if it is OS/driver issue of hardware

Re: Poor browser performance in OpenBSD

Have you tried vmstat to find bottleneck: is it CPU in userspace or in kernel or storage? Which WM are you using? Could it be WM issue? Try to use lower resolution (I know that resolution is locked for LCDs, but you can try just to make sure it is not the problem)

Re: Web store

AFAIK ACME (used by acme-client(1)) is supported by letsencrypt , so you do not ever need certbot: https://www.vultr.com/docs/using-let-s-encrypt-on-openbsd-6-1 Seems that everything should work with base system only (I have not tried though) On Wed, Jun 13, 2018 at 2:37 AM, Daniel Corbe

Re: [OT] VirtualBox guest additions support for OpenBSD

I am also interested in this feature, but here is lifehack I use to share clipboard and workaround poor desktop experience of vesa(4): I run X server on my host machine, and redirect output to it using DISPLAY env var. So, all apps are running on OpenBSD/VBox but with full screen and shared

Re: USB power management

I am not 100% sure, but I believe that to disable USB device OS uses "suspend" device hub controller feature. According to http://openbsd-archive.7691.n7.nabble.com/Suspending-USB-devices-from-userland-td259075.html This is not possible in OpenBSD. On Tue, Jun 12, 2018 at 9:22 PM, Jeff Ross

Re: Web store

..And certificate is still broken for chrome:( On Tue, Jun 12, 2018 at 7:08 PM, Base Pr1me wrote: > Who runs https://www.openbsdstore.com? I went to buy a couple of shirts > last Friday, but cert returns errors and paypal linking stuff is quite > broken. Also, no one is responding to the

Re: Reboot loop

> OpenBSD doesn't use ACPI to find an isa UART, it only looks in the fixed > locations compiled in to the kernel. Ok, I see that "com.c" does it by reading register, it even has comment "Probe for all known forms of UART" > For a system console (with access to DDB etc.) you need a

Re: Reboot loop

There is > com0 at isa0 port 0x3f8/8 irq 4: ns16550a, 16 byte fifo in your dmesg. So, I assume your box reports com port somehow (via ACPI probably) Some boxes may have comport built into chipset but no external cable for it. I have one, I bought cable separately. Another option is to use UART

Re: Reboot loop

Ok, then try to follow Stuart Longland's advice: use serial console. Connect your PC using null-modem cable to another pc, and in boot(8) prompt type: boot> set tty com0 On another PC run cu(1) or minicom or screen (or for Windows you may use PuTTY), connect to OpenBSD and you will see all your

Re: Reboot loop

ddb(4): "ddb is invoked upon a kernel panic when the sysctl(8) ddb.panic is set to 1". I belive this value is default. So, kernel should be dropped into ddb on panic. Does it happen? What exactly do you see on screen along with uvm_fault? Do you see whole stacktrace? Check

Re: Reboot loop

https://www.openbsd.org/report.html See "How to create a problem report" step 5

Re: relayd(8) relay to unix domain socket

> > > That's a good question. A similar question I would have is whether it is > able to relay connections arbitrarily between IPv4 and IPv6. > from relayd.conf: inet6 address-prefix: If the requested destination is an IPv4 address, relayd(8) will forward

relayd(8) relay to unix domain socket

Hello, Is it possible to relay to unix domain socket using relayd(8)? It seems that relayd(8) only works with protocols on top of IP. Here is my usecase: I have application server that only able to listen HTTP. (there are alternatives that work with fastcgi, and I can use them with httpd(8), but

Re: OT: how do you write your tools /scripts for everyday tasks

> > > Or when the tool would be running long enough that the performance > difference matters. Also, Javascript/Perl/Python/Ruby/shell all tend > to be lousy at dealing with anything where control over timing is the > overriding issue. > > Or when your target environment needs you to be miserly

Re: OT: how do you write your tools /scripts for everyday tasks

Hello, Running external process from Python script is not good in most cases. It is better to use wrapper or binding. Search pypi.org for it. I use shell (ksh,sed,awk,tr,mail etc) for simple tools, and Python for something which is more complex. Python is used for scripting in many modern linux

Re: socket permissions inside /var/www chroot

Hello, I am definitely not an expert in this field, but here are some thoughts: connect to a fastcgi socket unless the socket's user and group are also > www:www. > Should not unix domain sockets be treated as regular files in case permissions? If yes, then httpd should be able to access any

Re: Building OpenBSD and ports VS installing from packages

>> do I still get the same package ? Yes. cc(1) does not use microarchitecture-specific features unless you provide "-march" explicitly. Other BSDs do it, but OpenBSD does not. So, cc(1) only knows that you are building something for amd64. There should not be any difference between nehalem and

Re: Building OpenBSD and ports VS installing from packages

Hello. OpenBSD team does not recommend to build anything that exists in packages. >>If so, building from ports would produce a different code? In most cases ports are not aware of your microarchitecture. See my question and Theo's answer.

Re: Viewport for man.openbsd.org -- readability on phones

> > > If it is not a secret, what runs behind man.openbsd.org? Like httpd, CGI? > According to response headers: "Server: OpenBSD httpd". And with httpd(8) it must be FastCGI implemented either by perl script directly or with aid of slowcgi(8)

Re: print usb printer by [ Google Cloud Print for Chromium ]

Hello Jordan, >> you can do some neat things and avoid having to remove the ulpt(4) driver from the kernel What can be the reason to remove it? If I understand it correctly, this driver provides support for "USB printer class" (

Re: OpenBSD 6.3 AMD63 ond Dell Vostro 5470 System BIOS problems

Are you using BIOS/MBR or UEFI/GPT? It latter case, try BIOS/MBR.

Re: Remote kernel debugging with kgdb and vmm

AFAIK kgdb (remote debug) is not supported: http://openbsd-archive.7691.n7.nabble.com/on-line-kernel- debugging-tt335833.html#a341551 You can debug running kernel but functionality is limited.

Re: CVE-2018-8897

> > >> Then how do they implement memory watch? >> > > Got me, but even the ancient, in-tree gdb is able to do so. Have you > consulted the gdb source? > I read gdb sources and found an asnwer, but later I read docs and here it is: https://sourceware.org/gdb/onlinedocs/gdb/Set-Watchpoints.html

Re: CVE-2018-8897

Hello Mike, >> OpenBSD does not allow userspace to access the hardware debug registers. Does it mean that gdb and other debuggers can't use hardware breakpoints? Then how do they implement memory watch? AFAIK in other OSes they modify DR* registers using ptrace(2), but "struct reg” from used

Re: Can SSH report successful connections to pf?

If you want to open gate for those, who authenticated using ssh, you may need authpf(8) (known as Authentication Gateway) https://www.openbsd.org/faq/pf/authpf.html

Re: Best Practices python virtualenv

> while there are Python modules which rely on W|X. Yes, but I do not use them. I only run Python and Django. I am aware of the fact that my python is not compatible with some modules, and I am ok with it since I do not need them. > use ports instead of pkg_add to get Python binary without W|X

Re: Best Practices python virtualenv

> > So you're now on a custom built python and are unable to use standard pkg_add upgrades to new versions. I'd say that on balance, this is more > likely to *reduce* your security. I built it from ports, so I can update it using standard port update procedure. Since packages are built from

Re: Best Practices python virtualenv

The only difference is venv creates link to python, it does not copy binary itself. You now have python3 -> /usr/local/bin/python3 in your venv. Since /usr/local/ has wxallowed by default (see your /etc/fstab) it works. Does it affect security? In theory -- yes, because python can now create WX

Re: Best Practices python virtualenv

It is up to you, but I still belive that best solution is to rebuild python without of wxneeded. 1) It improves security 2) It fixes your virtualenv issue. If you do not use packages that need WX, why do you need wxneed?

Re: Best Practices python virtualenv

Hello. Short answer: if you do not need py-cryptography and QtWebKit, just rebuild python from ports disabling USE_WXNEEDED. I run Django using virtualenv in my $HOME and it works. Long answer: To use mmap(2) with PROT_EXEC | PROT_WRITE you need to link binary with -z wxneeded (See ld(1)). It

Re: Disk disappeared from system after cold reboot .

Hello. I am not OVH expert, but from dmesg I can't see second drive. I see only sd0 (20480MB) connected to your VirtIO. Dobule check that both drives are connected. > why is it mentioning GUID and not disk path ? It can work with drive names also, i.e: "/dev/sd0a /" but name may be changed

Re: Clarification re: rebuilding softraid mirror

Hello, No, you do not need to reboot. At least this is how it worked for me for raid 1: 1) bioctl softraid0 said raid degraded 2) I installed new disk (sd2). 3) kenrel reported on console that disk is detected 4) I created MBR using fdisk on it 5) I created disklabel with RAID type on it 6)

Re: Virtualbox vs latest snapshot

Latest snapshot running on 5.2.8 @win7x64 (everything works and seems to be stable) Do you use GPT or MBR? Is it SATA or PATA/IDE? Can you boot your hda0:/bsd from cd? Can you boot from CD, mount drive and fsck it? xx:2$ dmesg OpenBSD 6.3-current (GENERIC) #13: Thu Apr 26 07:46:22 MDT 2018

Re: PERC 6/i existence of Virtual Drives hangs boot

>> Or, is there a gdb method of booting the kernel? Short answer: http://openbsd-archive.7691.n7.nabble.com/on-line-kernel-debugging-td335833.html Long answer: /dev/kmem (kmem(4)) could be used to access kernel memory. gdb has special target (kvm) that uses it to debug bsd kernels (

Re: NFS keeps crashing

> I use it on osx: it is a crippled version of the original that causes endless problems with file permissions. Rock solid for the rest. File permissions are always hard to maintain between Windows and *nix, because of Windows ACL's and posix permissions differences. One person on this list

Re: 6.3/amd64 Thinkpad T530 touchpad problem (was ok in 6.2/amd64)

+1 for trying synaptics(4). It uses wsmouse "absolute mode" (covered by wsmouse(4)) and has a lot of Options (more than ws(4) which is used by default), some of them may help. Check log file then, you should see "synaptics" driver there. You can also try to stop wsmoused(8) before launching X.

Re: NFS keeps crashing

> I mean sponsors who pay for projects and compatibility updates. I also mean broader user base. IMHO: To increase userbase and attract serious sponsors attention, OpenBSD needs to be converted to "friendly-for-non-IT-people" OS like Ubuntu. Do you want to have polkit and Systemd in base

Re: Boot problem after power failure in OpenBSD 6.2 and later versions

> > Reinstalling the operating system seems to solve the problem. > Almost never you need to reinstall OpenBSD. There are only 2 parts that could be broken in your case: boot(8) and kernel itself (/bsd). Both could be downloaded from CD or ftp.OpenBSD.org website That is why I told you to try to

Re: Boot problem after power failure in OpenBSD 6.2 and later versions

in "boot -s", the device reboots itself again. > As you said, I booted from flash / cdrom and fsck all the partitions, but > that did not help either. > Also I checked /bsd exists on root > However, the system can not be booted. > > I dont know what is the weakness. > T

Re: Boot problem after power failure in OpenBSD 6.2 and later versions

. Are they clean? Mount them, and check /bsd exists on root (.a) On Fri, Apr 20, 2018 at 11:23 PM, augusta bonaventura <augusta...@gmail.com> wrote: > It means OpenBSD is constantly restarting itself. > > 2018-04-20 23:01 GMT+03:00 IL Ka <kazakevichi...@gmail.com>: > >

Re: Boot problem after power failure in OpenBSD 6.2 and later versions

> When the device reboots, it reboots itself when > it comes to the "boot>" menu. What do you mean "reboots itself "? boot(8) reboots your machine instead of booting kernel with out of any output?

Re: OpenBSD + 3G/4G USB modem

Do you have ucom at umodem in dmesg? Or ucom at umsm? On Fri, Apr 20, 2018 at 9:17 PM, MS wrote: > I forgot to mention it but the modem doesn't respond on any of the > /dev/cuaUX > > 2018-04-20 11:17 GMT+02:00 Roderick : > > > > > > > On Fri, 20 Apr

Re: How to maintain/debug OpenBSD

not need anything except man pages for tools. On Fri, Apr 20, 2018 at 6:35 PM, <sunny.na...@igreen.pk> wrote: > Is there any books that describe how to deal with these tools ? > > Sunny Naqvi > > On Fri, Apr 20, 2018 at 04:42 PM, IL Ka <kazakevichi...@gmail.com> wro

Re: vmstat

vmstat can't be used for that. Use top(1) and check its SIZE (virtual mem size) and RES (mem currently in RAM) columns. You can also use ps(1) with -m argument to sort by memory usage. It has %MEM (compared to other processes), RSS (like RES), VSZ (like SIZE) read man pages for ps and top:) On

Re: How to maintain/debug OpenBSD

> high load top, vmstat. You can also try to profile app with dtrace. > low space disk ``du -d1 -h`` to find huge dir > see opened files fstat On Fri, Apr 20, 2018 at 9:56 AM, wrote: > Hi > > Sometimes we need to maintain or debug OpenBSD. > I found these links

Re: sshfuse: fusefs: libfuse vnode reclaim failed

r other short term, uninterruptible (!)) wait. See also https://www.novell.com/support/kb/doc.php?id=7002725 On Fri, Apr 20, 2018 at 9:40 AM, Rudolf Sykora <rudolf.syk...@gmail.com> wrote: > Dear IL, > > On 20 April 2018 at 02:34, IL Ka <kazakevichi...@gmail.com> wr

Re: sshfuse: fusefs: libfuse vnode reclaim failed

First, make sure your SSH connection is stable. Try to SSH to your server, run some commands, download big (several megabytes) file using SCP or SFTP. It could be that your SSH problems. If it works, then we can try to investigate it: > libfuse vnode reclaim failed This message is from

Re: kernel relink segfaults on ALIX

> commented out the call to > reorder_kernel in rc. I still believe that ``chmod -x /usr/libexec/reorder_kernel`` is much better and safer than patching ``/etc/rc``, but it is up to you)

Re: kernel relink segfaults on ALIX

I am sorry, I skipped your last line: > Just comment in rc -_- for kernel You are right. But it is safer to "chmod -x" reorder_kernel (touching rc is not good) On Thu, Apr 19, 2018 at 9:58 PM, sven falempin <sven.falem...@gmail.com> wrote: > On Thu, Apr 19, 2018 at 1:01

Re: kernel relink segfaults on ALIX

manner). We can disable library_aslr, but there is no same option for kernel.. On Thu, Apr 19, 2018 at 9:58 PM, sven falempin <sven.falem...@gmail.com> wrote: > On Thu, Apr 19, 2018 at 1:01 PM, IL Ka <kazakevichi...@gmail.com> wrote: > > > Upgrade may affect kernel

Re: kernel relink segfaults on ALIX

rvives an > upgrade. > On Apr 19, 2018 9:44 AM, IL Ka <kazakevichi...@gmail.com> wrote: > > > > Ancient UNIX way to disable anything: ``doas chmod -x > > /usr/libexec/reorder_kernel`` ;) > > > > Although ``reorder_kernel`` is very simple ksh script, I agree it should > be > > documented. > > >

Re: 6.3/amd64 Thinkpad T530 touchpad problem (was ok in 6.2/amd64)

I forgot to mention that you can disable extended input devices. Ie. $ xinput --list Then look at list of "Virtual core XTEST pointer" children And ``xinput --disable [id]`` everything but the first one. It _may_ help. On Thu, Apr 19, 2018 at 2:38 PM, Jonathan Thornburg <

Re: 6.3/amd64 Thinkpad T530 touchpad problem (was ok in 6.2/amd64)

Try to start ``wsmoused(8)`` and check if mouse works in console. ``/etc/rc.d/wsmoused start`` and move mouse around for minute or two. Does it work? It will help us to understand if it is a X problem or wmouse(4) problem

Re: kernel relink segfaults on ALIX

Ancient UNIX way to disable anything: ``doas chmod -x /usr/libexec/reorder_kernel`` ;) Although ``reorder_kernel`` is very simple ksh script, I agree it should be documented. On Thu, Apr 19, 2018 at 12:15 PM, Z Ero wrote: > Coincidently I just logged in to write the

Android (MTP) with OpenBSD: Tiny success story

I just connected my Android device to OpenBSD, and since I did not find any article on this subject, I want to share my experience. OpenBSD supports USB Mass Storage Device (used in usb drives) with umass(4). But Android uses MTP (file-level protocol, not block-level like umass), So OpenBSD

No place for "wsfontload" and "wsconscfg" in rc?

Hello, It may sound silly: nobody use console these days (except emergency), but I am curious: I want to load font and set it using wsconsctl display.font I also want to change display type. I can do all of that in rc.local, but there is a separeate place for wsconsctl : /etc/wsconsctl.conf I

Re: NFS server down, again, and again, and again...

You could use ktrace(1) to trace all calls and then use kdump(1) to read them, and may help you to find what cause it to die, but it may be tricky for anyone except nfsd developer.. You can also try to find person who supports it by looking at last commits to:

Re: NFS server down, again, and again, and again...

Do you mean nfsd server dies? Does it provide core dump? You do not need to restart it manually: just create script that checks for server existence (like ``/etc/rc.d/nfsd check``) and run it if it is dead. On Wed, Apr 18, 2018 at 12:00 AM, Rupert Gallagher wrote: > The

Re: Beg for Atheros wifi driver

> 0x: Vendor ID: 168c Product ID: 003e https://vendev.org/pci/ven_168c_003e/ "QCA6174 802.11ac Wireless Network Adapter" You can try $ man pci | grep Atheros or $ apropos Atheros To get list all Atheros-related drivers ( ath(4), athn(4) and same for usb) but it seems that they do not

Re: trouble installing php on 6.3

2018 at 5:28 PM, Michael Maurer <maurer.mich...@gmail.com> wrote: > On 17 April 2018 at 14:13, IL Ka <kazakevichi...@gmail.com> wrote: > > So, does it work only with PKG_TMPDIR set? > > > > # PKG_TMPDIR=/tmp/foo pkg_add -r -vvv curl > > (-r is used to replac

Re: trouble installing php on 6.3

So, does it work only with PKG_TMPDIR set? # PKG_TMPDIR=/tmp/foo pkg_add -r -vvv curl (-r is used to replace old package) Could it be your harddrive or filesystem problem? Try to fsck partition where pkg_add stores packages. If you are able to download file by wget, you can install it using

CPUTYPE in mk.conf to set -march like in FreeBSD?

In FreeBSD they have /etc/make.conf You set CPUTYPE= there, and /usr/share/mk/bsd.cpu.mk will read it, and set -march for CPUCFLAGS This variable is used by ports, kernel, and any random Makefile may use it. (There are similar things in Linux Gentoo also) But OpenBSD /etc/mk.conf does not have

Re: Beg for Atheros wifi driver

Almost all drivers in OpenBSD report themselves, so you can grep your dmesg for message like "foo0 at pci0 dev0 function 0 [PCI Device name from PCI configuration space] " $ man autoconf $ man dmesg Except dmesg, this information is reported to userspace via /dev/pci and could be read with

Re: DMESG / syslog

Hello. This is "kern" facility. # touch /var/log/dmesg And then add "kern.* /var/log/dmesg" to your "syslog.conf". Restart syslog and anything printed by kernel will go to /var/log/dmesg But note that when booting, kernel does not have access to syslog, so it stores messages in its internal

Re: trouble installing php on 6.3

Hm.. sounds strange then. Did you try to call "pkg_add libxml"? If not, then try and in case of success try pkg_add php again If it does not work, then lets do the following: To make sure libxml is deleted # pkg_delete libxml use different mirror # echo

Re: Community-driven OpenBSD tutorials wiki?

woman (with out man) is tool used in Emacs to read manual pages with out of actually running man https://www.gnu.org/software/emacs/manual/html_mono/woman.html :) On Mon, Apr 16, 2018 at 6:25 PM, Mehma Sarja wrote: > It is meant as a play on words, a light hearted

Re: trouble installing php on 6.3

php depends on libxml, pkg_add tries to download it, and downloaded archive is broken. pkg_add uses "/etc/installurl" to find website to download packages. You may try different website (echo "https://ftp.openbsd.org/pub/OpenBSD/; > /etc/installurl) Also, try to download this file and unpack

Re: OpenBSD 6.3 kernel panic

Technically, you can debug your real machine kernel using serial (aka com) port, but I do not think you need to do that. When your kernel panics, it takes your to ddb, right? You got "uvm_fault page fault trap": it means kernel tries to access some invalid memory address. It happens in

Re: vgafb manpage is in wrong place, should I report it?

> When a manual page applies to more than 1 arch, we place it a level up. Thank you, but should not we add comment like the one we have for "man pci" SYNOPSIS # macppc, sparc64 vgafb* at pci? Btw, I found some more things to improve in mans like: options(4) APERTURE section should have

Re: OpenBSD 6.3 kernel panic

> > Maybe kernel panic isn't the correct terminology? ddb is invoked anytime I > touch the track pad while booting. You are right: it is kernel panic. Accroding to ddb(4): "ddb is invoked upon a kernel panic when the sysctl(8) ddb.panic is set to 1." I am not sure, but I believe you can

vgafb manpage is in wrong place, should I report it?

Hello, I am pretty new to OpenBSD and not sure if I should report minor issue to bugs@, but I just found that vgafb(4) device is supported only on macppc and sparc64 (according to src/sys). It has nothing to do with i386 nor with amd64. But for some reason it's man page sits directly in

Re: Xfce with the .xinitrc file

Hi, So, does xfce4-session start with out of console kit or does not start at all? Do you use display manager or startx? Note that .xinitrc is used only by xinit (startx), not by xenodm/xdm. ck-launch-session is part of ConsoleKit, do you have it installed? On Fri, Apr 13, 2018 at 1:38 PM,