Re: CVS commit: src/sys

2012-05-31 Thread David Laight 
On Wed, May 30, 2012 at 05:11:09PM -0700, John Nemeth wrote:
 On Oct 11,  4:10pm, David Laight wrote:
 } 
 } Module Name:src
 } Committed By:   dsl
 } Date:   Mon May 21 21:34:16 UTC 2012
 } 
 } Modified Files:
 } src/sys/arch/i386/stand/lib: exec.c
 } src/sys/arch/sandpoint/stand/altboot: main.c
 } src/sys/lib/libsa: ext2fs.c ffsv1.c ffsv2.c globals.c lfsv1.c lfsv2.c
 } stand.h ufs.c
 } 
 } Log Message:
 } Remove the code that tries to load the ffs kernel module during boot.
 } This is in line with the core decision than even modular kernels should
 }   contain the ffs code.
 } I've left in the code that tries to load nfs and ext2fs, but it
 }   isn't clear that is necessary.
 } Removes a warning message that (usually) flashes past to fast to read.
 } AFAICT all the relevant kernels contain ffs (and nfs for that matter).
 
  Just because all shipping kernels are currently required to
 contain ffs doesn't mean a user can't build a kernel without it
 knowing that /boot will load the appropriate module.  At the very
 least, this should have been a knob so that people can build a /boot
 that will load the appropriate filesystem if they wish.

IIRC you can explicitly request (from boot.cfg) that any module be
loaded.  There is no need for boot itself to always try to load
such a module.

The problem is that there was no was of stopping boot trying to
load (typically) ffs.kmod, and if the kernel does contain it then
it doesn't seem correct at all.

David

-- 
David Laight: da...@l8s.co.uk

Re: CVS commit: src/tests/lib/libexecinfo

2012-05-31 Thread Martin Husemann 
On Wed, May 30, 2012 at 03:11:58PM +, Jukka Ruohonen wrote:
 Log Message:
 Use atf_arch instead of atf_machine; see atf-config(1).

What is the difference? atf-config(1) is mumbling about unnamed bugs
and upstream fixes, but does not explain what differs and which should
be prefered.

Martin

Re: CVS commit: src/tests/lib/libexecinfo

2012-05-31 Thread Jukka Ruohonen 
On Thu, May 31, 2012 at 12:29:55PM +0200, Martin Husemann wrote:
 On Wed, May 30, 2012 at 03:11:58PM +, Jukka Ruohonen wrote:
  Log Message:
  Use atf_arch instead of atf_machine; see atf-config(1).
 
 What is the difference?

I think there is no such thing as atf_machine.

- Jukka.

Re: CVS commit: src/tests/lib/libexecinfo

2012-05-31 Thread Martin Husemann 
On Thu, May 31, 2012 at 01:58:50PM +0300, Jukka Ruohonen wrote:
 I think there is no such thing as atf_machine.

Well, it worked for me ;-)

 atf_machine   The machine type name detected by ATF.  This should
   not be tunable but is provided for symmetry with
   atf_arch.

Martin

Re: CVS commit: src/tests/lib/libexecinfo

2012-05-31 Thread Jukka Ruohonen 
On Thu, May 31, 2012 at 01:00:52PM +0200, Martin Husemann wrote:
 Well, it worked for me ;-)
 
  atf_machine   The machine type name detected by ATF.  This should
not be tunable but is provided for symmetry with
atf_arch.

My bad. It worked according to the logs too:

build: OK with 360028 lines of log, install: OK, tests: 3049 passed, 92
skipped, 40 expected_failure, 1 failed, ATF output: raw, xml, html

commit 2012.05.30.06.01.22 martin src/tests/lib/libexecinfo/t_backtrace.c 1.3
build: OK with 361963 lines of log, install: failed

skipped, 40 expected_failure, 0 failed, ATF output: raw, xml, html

http://releng.netbsd.org/b5reports/i386/commits-2012.05.html#end

- Jukka.

Re: CVS commit: src/sys

2012-05-31 Thread David Laight 
On Thu, May 31, 2012 at 10:06:28AM +0200, Martin Husemann wrote:
 On Thu, May 31, 2012 at 07:38:53AM +0100, David Laight wrote:
  IIRC you can explicitly request (from boot.cfg) that any module be
  loaded.  There is no need for boot itself to always try to load
  such a module.
 
 Yes there is, and I agree that this is the proper way to load it if you
 create a special ffs-as-modle-only kernel.

But you don't want /boot to try to load the module if the ffs code is
present in the kernel. Since /boot has no way of knowing what is in the
loaded kernel it is inappropriate for it to try to load the module file
and for the kernel to then try to load the module.

David

-- 
David Laight: da...@l8s.co.uk

Re: CVS commit: src/sys

2012-05-31 Thread Martin Husemann 
On Thu, May 31, 2012 at 06:30:38PM +0100, David Laight wrote:
 But you don't want /boot to try to load the module if the ffs code is
 present in the kernel. Since /boot has no way of knowing what is in the
 loaded kernel it is inappropriate for it to try to load the module file
 and for the kernel to then try to load the module.

I thought I agree with you - but now you confused me. If I create (with
care) a kernel w/o internal ffs, I can easily edit /boot.cfg. Hard
coding it in boot seems wrong. Isn't that what you said as well?

Martin

Re: CVS commit: src/sys

2012-05-31 Thread David Laight 
On Thu, May 31, 2012 at 08:32:42PM +0200, Martin Husemann wrote:
 On Thu, May 31, 2012 at 06:30:38PM +0100, David Laight wrote:
  But you don't want /boot to try to load the module if the ffs code is
  present in the kernel. Since /boot has no way of knowing what is in the
  loaded kernel it is inappropriate for it to try to load the module file
  and for the kernel to then try to load the module.
 
 I thought I agree with you - but now you confused me. If I create (with
 care) a kernel w/o internal ffs, I can easily edit /boot.cfg. Hard
 coding it in boot seems wrong. Isn't that what you said as well?

I think I misunderstood you...

David

-- 
David Laight: da...@l8s.co.uk