gmane.os.netbsd.general

Re: How (and wherther) to set pkg wpa_supplicant

Greg Troxel — 2013-06-19T16:21:54

Mayuresh <mayuresh< at >acm.org> writes:


Generally, one should be able to set a variable in
/etc/rc.conf.d/wpa_supplicant to make the change.  A real example:

  $ cat /etc/rc.conf.d/postfix 
  postfix_command="/usr/pkg/sbin/${name}"

But I'm not sure that works with wpa_supplicant; you'll have to read the
script.

How (and wherther) to set pkg wpa_supplicant

Mayuresh — 2013-06-19T15:55:16

I have been using wpa_supplicant from base and did not realize it has pkg
version as well, which is newer.

Would like to try using the pkg version to see whether it helps solve
latency issues (which I am discussing on a separate thread).

Particularly, how to set up /etc/rc.d/wpa_supplicant when using pkg
version?

(I just manually changed all the paths from /usr/sbin to /usr/pkg/sbin/ in
/etc/rc.d/wpa_supplicant. But not sure whether that is the right way to
go.)

Mayuresh.

Re: Frustrating network latency issue on NetBSD 6.1

Greg Troxel — 2013-06-19T15:53:42

Use ping, not ssh, to be able to have a more isolated test case and get
quantitative data.  In particular, separate loss from delay; both
present as delay over TCP.

Run tcpdump on both ends, and compare timestamps of packets seen in both
places, taking care to synchronize clocks (hard with network latency) or
be aware of the mis-synchronization.

Check to see if there are any power-saving settings for the adaptor and
the DD-WRT box and diable them.

Frustrating network latency issue on NetBSD 6.1

Mayuresh — 2013-06-19T15:49:29

I am having some terrible network latency issues with NetBSD 6.1 i386,
since my last hardware upgrade. Sadly, both the laptop and router are new
so not able to figure out what contributes to the issue.

One observation is, on the same setup, if I boot to Linux I do not find
the latency issue at all.

What I call latency issue basically appears when I connect to a remote (or
even on my home network) machine over ssh, typing is very jerky. A lot of
characters appear together in bunches after a few seconds of typing them.
It's getting unbearably frustrating.

Would appreciate help in understanding and resolving the issue.

lspci shows the wifi adapter as:
02:00.0 Network controller: Intel Corporation Centrino Advanced-N 6205
(rev 34)

It is configured as iwn0.

WPA2-PSK encryption is used as only that is supported on the router that
runs DD-WRT firmware.

(Ok, an obvious thing I'll try is by disabling encryption for a while.)

Mayuresh

Re: How (and wherther) to set pkg wpa_supplicant

Greg Troxel — 2013-06-19T16:21:54

Mayuresh <mayuresh< at >acm.org> writes:


Generally, one should be able to set a variable in
/etc/rc.conf.d/wpa_supplicant to make the change.  A real example:

  $ cat /etc/rc.conf.d/postfix 
  postfix_command="/usr/pkg/sbin/${name}"

But I'm not sure that works with wpa_supplicant; you'll have to read the
script.

How (and wherther) to set pkg wpa_supplicant

Mayuresh — 2013-06-19T15:55:16

I have been using wpa_supplicant from base and did not realize it has pkg
version as well, which is newer.

Would like to try using the pkg version to see whether it helps solve
latency issues (which I am discussing on a separate thread).

Particularly, how to set up /etc/rc.d/wpa_supplicant when using pkg
version?

(I just manually changed all the paths from /usr/sbin to /usr/pkg/sbin/ in
/etc/rc.d/wpa_supplicant. But not sure whether that is the right way to
go.)

Mayuresh.

Re: Frustrating network latency issue on NetBSD 6.1

Greg Troxel — 2013-06-19T15:53:42

Use ping, not ssh, to be able to have a more isolated test case and get
quantitative data.  In particular, separate loss from delay; both
present as delay over TCP.

Run tcpdump on both ends, and compare timestamps of packets seen in both
places, taking care to synchronize clocks (hard with network latency) or
be aware of the mis-synchronization.

Check to see if there are any power-saving settings for the adaptor and
the DD-WRT box and diable them.

Frustrating network latency issue on NetBSD 6.1

Mayuresh — 2013-06-19T15:49:29

I am having some terrible network latency issues with NetBSD 6.1 i386,
since my last hardware upgrade. Sadly, both the laptop and router are new
so not able to figure out what contributes to the issue.

One observation is, on the same setup, if I boot to Linux I do not find
the latency issue at all.

What I call latency issue basically appears when I connect to a remote (or
even on my home network) machine over ssh, typing is very jerky. A lot of
characters appear together in bunches after a few seconds of typing them.
It's getting unbearably frustrating.

Would appreciate help in understanding and resolving the issue.

lspci shows the wifi adapter as:
02:00.0 Network controller: Intel Corporation Centrino Advanced-N 6205
(rev 34)

It is configured as iwn0.

WPA2-PSK encryption is used as only that is supported on the router that
runs DD-WRT firmware.

(Ok, an obvious thing I'll try is by disabling encryption for a while.)

Mayuresh

Re: XBMC?

Pierre Pronchery — 2013-06-16T13:22:07

That's a fork of the upstream xbmc project - no connection to pkgsrc 
whatsoever. From there the course will typically be to:
- send pull requests (patches) to the upstream project;
- while they get accepted, backport and duplicate them to whichever
  version is found in pkgsrc.

Another alternative would be to package this repository instead of any 
official release upstream.


HTH,

Re: 6.1 upgrade solved audio problem, but espeak still doesn't work

Mayuresh — 2013-06-15T18:42:14

Correction. Dumping espeak output to wav file works if played with
"mplayer -ao sun". But mplayer is otherwise seen working fine with oss
since upgrading to 6.1.

Mayuresh.

6.1 upgrade solved audio problem, but espeak still doesn't work

Mayuresh — 2013-06-15T09:34:17

On 60 i386, intel audio hardware as described in the following thread,
audio wasn't working on 6.0.

http://mail-index.netbsd.org/netbsd-users/2013/05/11/msg012792.html

After upgrading to 6.1, mplayer has started working, with default oss
device, but espeak text to speech converter still doesn't work.

I don't find options in espeak to override audio driver etc.

Dumping espeak output to wav file and playing with mplayer doesn't work.
There is no sound produced.

Redirecting espeak output to stdout and in turn to /dev/audio0 only
produces some noise.

Any leads to get espeak work?

espeak worked on my previous hardware with 6.0/i386 and an older version
of intel audio hardware. (Don't have details of old audio hardware now.)

Mayuresh

Re: Synaptics Touchpads

Christos Zoulas — 2013-06-15T00:08:52

In article <DUB119-W48786D1653B6C7A8FCB21581800< at >phx.gbl>,
Graham Sparks  <gmjs< at >hotmail.co.uk> wrote:

I would turn on some debugging. Or put some printfs in the code.

christos

Synaptics Touchpads

Graham Sparks — 2013-06-14T11:08:03

Hello,

I'm new to BSD and am trying to instal NetBSD (6.1 amd64) on a laptop, but if the touchpad is pressed (console or X session), both the keyboard and touchpad stop responding.

A USB keyboard and mouse connected work fine.

The touchpad appears to be detected correctly (dmesg output):

pms0 at pckbc1 (aux slot)
pms0: Synaptics touchpad version 7.2
pms0:Palm detect
pckbc1: using irq 12 for aux slot
wsmouse0 at pms0 mux 0


I'm using the GENERIC kernel, which appears to enable support for touchpads by default:

# Console Devices
# wscons
pckbc0        at isa?            # pc keyboard controller
pckbd*        at pckbc?        # PC keyboard
pms*        at pckbc?        # PS/2 mouse for wsmouse
#options     PMS_DISABLE_POWERHOOK    # Disable PS/2 reset on resume
options     PMS_SYNAPTICS_TOUCHPAD    # Enable support for Synaptics Touchpads
options     PMS_ELANTECH_TOUCHPAD    # Enable support for Elantech Touchpads
vga*        at pci? dev ? func

Re: libXinerama.so.2 - where to find this one?

herbert langhans — 2013-06-11T20:25:31

    From: Manuel Bouyer <bouyer< at >antioche.eu.org> [130611 21:46]
    On Sun, Jun 09, 2013 at 06:01:40PM +0200, herbert langhans wrote:
    > Ok - now I tried to make a symlink from the existing libXinerama.so.1 to
    > the supposed-to-be libXinerama.so.2 - and then the existing libXi.so.6
    > to libXi.so.7 and so on.
    > 
    > It's the XPCOMGlueLoad error - but what package supposed to contain the
    > XPCOMGlueLoad-whatever files??
    > 
    > XPCOMGlueLoad error for file /usr/pkg/lib/seamonkey/libxpcom.so:
    > Shared object "libXi.so.7" not found
    > Couldn't load XPCOM.
    > ocelot$ seamonkey
    > XPCOMGlueLoad error for file /usr/pkg/lib/seamonkey/libxpcom.so:
    > Shared object "libXcursor.so.2" not found
    > Couldn't load XPCOM.
    > ocelot$ seamonkey
    > XPCOMGlueLoad error for file /usr/pkg/lib/seamonkey/libxpcom.so:
    > Shared object "libXcomposite.so.2" not found
    > Couldn't load XPCOM.
    
    Looks like you have serious packages mismatch.
    What NetBSD version are you r

Re: libXinerama.so.2 - where to find this one?

Manuel Bouyer — 2013-06-11T19:46:22

Looks like you have serious packages mismatch.
What NetBSD version are you running, and from where did you download
the packages ?

Re: addressing > 2 TB storage

Jeff Rizzo — 2013-06-11T16:08:31

There is some boot-from-gpt support;  I'm successfully doing it on one 
machine where all the disks are GPT.

See the "biosboot" option for gpt(8), and this (very terse) wiki page:

http://wiki.netbsd.org/users/jakllsch/gptboot/

+j

Re: addressing > 2 TB storage

Jeff Rizzo — 2013-06-11T16:29:52

I *think* so - the two or three times I've used it successfully (all but 
one under VMWare :), I will admit to not 100% understanding the details.

I will try to spend some time on boot(8) in the nearish future.

+j

Re: addressing > 2 TB storage

Greg Troxel — 2013-06-11T16:26:49

Jeff Rizzo <riz< at >tastylime.net> writes:


Cool - glad I was confused.

It would be great to get this into boot(8).

I gather that gptmbr.bin merely reads the GPT label to find the active
gpt partition and then reads the 2nd-stage bootxx_ from it.
So one uses gptmbr.bin instead of mbr_bootsel, essentially?

Re: addressing > 2 TB storage

Greg Troxel — 2013-06-10T19:57:50

cal lawton <cal< at >iglou.com> writes:


Basically, see gpt(8).   Using gpt, you can have partitions larger than
2T.   However, there is not as far as I know support for booting from
GPT.

Re: Mysql server NetBSD 6.1 RC4 it does not works

Franck Lesage — 2013-06-10T17:14:47

I installed NetBSD 6.1 RC4 since few times. I wanted to install mysql.
And I faced also a problem. First, I install mysql with pkgin:

pkgin in mysql-server-5.5.30

Then I try to initialize the mysql database:

cd /usr/pkg/share/mysql && cp my-medium.cnf my-default.cnf
mysql_install_db

The output of mysql_install_db for me is:

Installing MySQL system tables...130610 16:40:45 InnoDB: The InnoDB memory 
heap is disabled
130610 16:40:45 InnoDB: Mutexes and rw_locks use GCC atomic builtins
130610 16:40:45 InnoDB: Compressed tables use zlib 1.2.3
130610 16:40:45 InnoDB: Initializing buffer pool, size = 128.0M
130610 16:40:45 InnoDB: Completed initialization of buffer pool
130610 16:40:45 InnoDB: highest supported file format is Barracuda.
130610 16:40:45  InnoDB: Waiting for the background threads to start
130610 16:40:46 InnoDB: 5.5.30 started; log sequence number 1595675
ERROR: 1064  You have an error in your SQL syntax; check the manual that
corresponds to your MySQL server version for the right syntax to u

addressing > 2 TB storage

cal lawton — 2013-06-10T16:47:02

Imagine my surpise when I discovered that I can't see more than two terabytes….well 1.9 TB really.

My google-fu is usually strong in solving things like this, so I hope someone from the community has a technique, because I'm not seeing it.

The hardware setup is simple: esata system from SansDigital set to R5. The subsystem is working as my Mac can partition to 12 TB without error, same for MS Windows.


--
cal lawton
lexington, kentucky
cal< at >iglou.com