Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 14 Mar 2010 17:00:15 +0100 (CET)
From:      Juergen Lock <nox@jelal.kn-bremen.de>
To:        scdbackup@gmx.net
Cc:        freebsd-hackers@freebsd.org
Subject:   Re: How to slow down SATA to 1.5 GBit/s ?
Message-ID:  <201003141600.o2EG0F2M005027@triton8.kn-bremen.de>
In-Reply-To: <106070878026016@192.168.2.69>
References:  <201003132009.o2DK9nBg061996@triton8.kn-bremen.de>

next in thread | previous in thread | raw e-mail | index | archive | help
In article <106070878026016@192.168.2.69> you write:
>Hi,
Hi!
>
>Gary Jennejohn wrote:
>> I'm using ahci with two ATA/IDE DVD drives and they work as long as I
>> have "device ata" in my kernel config.  That's all I need.
>
>So i will probably need that for the IDE DVD-ROM
>drive that is built in too.
>
 Yeah I think so (I don't have ide drives on this box anymore...)
There is code in head now and (I think) stabe/8 to use cam for `regular'
(i.e. non-ahci) ata devices too, but I haven't played with that yet as
ahci(4) now works pretty well for me (this box is on stable/8.)
>
>Juergen Lock wrote:
>> thanks for trying to help improving libburn! :)
>
>Well, i am its active developer since a while.
>Any problems with it are probably my fault.
>
 :)
 
>Nevertheless, its FreeBSD SCSI transport is not
>originally by me. So i lack specific knowledge.
>(There are two alternative transport adapters:
>sg-freebsd originally by Alexander Nedotsukov,
>sg-libcdio for upcomming libcdio-0.83 .)
>
>
>> http://www.freebsd.org/cgi/query-pr.cgi?pr=138789
>
>I will test.
>This was BD-R, not BD-RE, i assume ?
>
 That 15GB disc was BD-RE actually.

>A leadout track. Sounds very CD-ish.
>With DVD and BD one should rather go for READ
>DISC INFORMATION and READ TRACK INFORMATION.
>
 Hmm you might want to followup on the PR with that hint...
>
>> Well if you didn't create a swap partition during install then you
>> don't have one. :)
>
>Not to have one seems to be an advantage now.
>I don't remember to have been asked.

 Well you'd have created it in the label editor, I don't think
sysinstall specifically asks for it.

> But with
>4 GB of RAM i might simply have answered 'no'.
>
 :)
>
>> the saved state of a panic'd kernel,
>
>I assume for reporting my kernel panics i have to
>install a development kernel ?

 Actually I _think_ GENERIC gets built with debug info by default
these days, look for /boot/kernel/*.symbols files to check.  Of course
it may still make sense to build a custom kernel with ddb enabled etc
too, see:
	http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/kernelconfig-building.html

 Here is a config you could use:

------------snip--------
include		GENERIC

ident		MYKERNEL
nooptions	WITNESS

#options 	WITNESS
#
# The INVARIANTS option is used in a number of source files to enable
# extra sanity checking of internal structures.  This support is not
# enabled by default because of the extra time it would take to check
# for these conditions, which can only occur as a result of
# programming errors.
#
#options 	INVARIANTS

#
# The INVARIANT_SUPPORT option makes us compile in support for
# verifying some of the internal structures.  It is a prerequisite for
# 'INVARIANTS', as enabling 'INVARIANTS' will make these functions be
# called.  The intent is that you can set 'INVARIANTS' for single
# source files (by changing the source file or specifying it on the
# command line) if you have 'INVARIANT_SUPPORT' enabled.  Also, if you
# wish to build a kernel module with 'INVARIANTS', then adding
# 'INVARIANT_SUPPORT' to your kernel will provide all the necessary
# infrastructure without the added overhead.
#
#options 	INVARIANT_SUPPORT
#
#
# Compile with kernel debugger related code.
#
options 	KDB

#
# Print a stack trace of the current thread on the console for a panic.
#
options 	KDB_TRACE

# Don't enter the debugger for a panic. Intended for unattended operation
# where you may want to enter the debugger from the console, but still want
# the machine to recover from a panic.
#
options 	KDB_UNATTENDED
#
# Enable the ddb debugger backend.
#
options 	DDB
#
# Enable the remote gdb debugger backend.
#
options 	GDB

# Size of the kernel message buffer.  Should be N * pagesize.
options 	MSGBUF_SIZE=(40960*4)
------------snip--------

>(Switching off-and-on a stuck USB drive is
> obviously not a healthy thing to do.)
>
 Hehe.

>Well, i plan to install another FreeBSD from
>scratch to try reproducing drive concurrency
>problems which i experienced first and which
>vanished after installing the ports of xfburn
>and xorriso.
>First drives got stuck when disturbed while
>burning CD. Now the offender gets blocked until
>the vulnerable drive state ends.
>Those ports triggered a cascade of other ports.
>So i have no idea which one might have tweaked
>the configuration (or whathever happened).
>
 Hmm not sure what happened there...

 Oh I do remember one issue:  Burning usually doesn't work when hald
is running, I suspect burning tools would have to tell it to stop polling
the drive at least while they are writing a disc...
>
>> Btw there also is siis(4) [...] for
>> SiliconImage sata controllers.
>> [...] sas controllers
>
>If somebody has such hardware then i would be
>interested to hear whether it works with libburn.
>Anybody is invited to ask for support.
>
 Actually I do have siis(4) here too, will have to test that someday...
>
>Have a nice day :)

 Ditto! :)
	Juergen



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201003141600.o2EG0F2M005027>