Peter Teoh | 1 May 01:35 2008
Picon

Re: arch/um/Makefile:100: *** missing separator. Stop.

On Thu, May 1, 2008 at 2:47 AM, Jeff Dike <jdike <at> addtoit.com> wrote:
> On Thu, May 01, 2008 at 12:24:21AM +0800, Peter Teoh wrote:
>  > Mine is same:
>  >
>  > md5sum arch/um/Makefile
>  > 9088cdab1c0b725568e8f269636dc6df  arch/um/Makefile
>
>  What make are you using?
>

Thank you Jeff,

/mnt/hd0/download/linux-2.6-latest>make -v
GNU Make 3.81
Copyright (C) 2006  Free Software Foundation, Inc.
This is free software; see the source for copying conditions.
There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
PARTICULAR PURPOSE.

This program built for i686-redhat-linux-gnu
/mnt/hd0/download/linux-2.6-latest>which make
/usr/bin/make

Any ideas?   As far as i can remember I don't think I have install any
"make".   And btw, I have three machine - FC5, FC6, FC7 and all have
the same symptom.   Steps taken are make mrproper, then make oldconfig
ARCH=um (or defconfig tried).   All three machines are running 2.6.25
(different rc version).

Please help ????
(Continue reading)

Jeff Dike | 1 May 17:13 2008

Re: actually useful backtrace from a CPU-chewing hang

On Wed, Apr 30, 2008 at 10:49:27PM +0100, Nix wrote:
> Done, and tested by stepping the time (five seconds --- five seconds per
> day! I have pendulum clocks that keep better time than that!), and,
> oops, instant loop as before, with this debugging patch applied, which
> did not fire...

Hmmm.  I stepped the time (date -set "now+5sec") and nothing bad
happened.  Is that more or less what you did?

Can you send me your config?

    	     	Jeff

--

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
Jeff Dike | 1 May 17:15 2008

Re: arch/um/Makefile:100: *** missing separator. Stop.

On Thu, May 01, 2008 at 07:35:32AM +0800, Peter Teoh wrote:
> /mnt/hd0/download/linux-2.6-latest>make -v
> GNU Make 3.81
> Copyright (C) 2006  Free Software Foundation, Inc.
> This is free software; see the source for copying conditions.
> There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
> PARTICULAR PURPOSE.
> 
> This program built for i686-redhat-linux-gnu
> /mnt/hd0/download/linux-2.6-latest>which make
> /usr/bin/make
> 
> Any ideas?   

No.  Something must be odd on your end.  Can you script a session
where you download the source and make this happen, then post it?  I'd
like to see a completely from-scratch session.

      	  	       Jeff

--

-- 
Work email - jdike at linux dot intel dot com

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
Peter Teoh | 1 May 18:13 2008
Picon

Re: arch/um/Makefile:100: *** missing separator. Stop.

On Thu, May 1, 2008 at 11:15 PM, Jeff Dike <jdike <at> addtoit.com> wrote:
> On Thu, May 01, 2008 at 07:35:32AM +0800, Peter Teoh wrote:
>  > /mnt/hd0/download/linux-2.6-latest>make -v
>  > GNU Make 3.81
>  > Copyright (C) 2006  Free Software Foundation, Inc.
>  > This is free software; see the source for copying conditions.
>  > There is NO warranty; not even for MERCHANTABILITY or FITNESS FOR A
>  > PARTICULAR PURPOSE.
>  >
>  > This program built for i686-redhat-linux-gnu
>  > /mnt/hd0/download/linux-2.6-latest>which make
>  > /usr/bin/make
>  >
>  > Any ideas?
>
>  No.  Something must be odd on your end.  Can you script a session
>  where you download the source and make this happen, then post it?  I'd
>  like to see a completely from-scratch session.
>
Ok, I tried on many different tree - linus tree, sched-devel, vfs tree
etc, and the linus tree repeated on 3 different machines - all having
the same problem as above.

Then I tried a specific stable version - 2.6.24.2.tar.bz2 as downloaded:

Filesize:
-rw-r--r-- 1 root root 46735221 Feb 15 23:27 ../linux-2.6.24.2.tar.bz2

md5sum ../linux-2.6.24.2*bz2
dd573a2fae55624ed92fa49c17583964  ../linux-2.6.24.2.tar.bz2
(Continue reading)

Peter Teoh | 1 May 18:18 2008
Picon

Re: arch/um/Makefile:100: *** missing separator. Stop.

In case if anyone is wondering what this line 107 means in the arch/um/Makefile:

    104 ifneq ($(KBUILD_SRC),)
    105 $(shell mkdir -p $(ARCH_DIR) && ln -fsn
$(srctree)/$(ARCH_DIR)/Kconfig.$(SUBARCH) $(ARCH_DIR)/Kconfig.arch)
    106 else
    107 $(shell cd $(ARCH_DIR) && ln -sf Kconfig.$(SUBARCH) Kconfig.arch)
    108 endif

Different version of Linux kernel tried, will always end up at the
same line above, btw, although the line number can be different.

-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
Nix | 2 May 01:34 2008
Picon

Re: actually useful backtrace from a CPU-chewing hang

On 1 May 2008, Jeff Dike outgrape:

> On Wed, Apr 30, 2008 at 10:49:27PM +0100, Nix wrote:
>> Done, and tested by stepping the time (five seconds --- five seconds per
>> day! I have pendulum clocks that keep better time than that!), and,
>> oops, instant loop as before, with this debugging patch applied, which
>> did not fire...
>
> Hmmm.  I stepped the time (date -set "now+5sec") and nothing bad
> happened.  Is that more or less what you did?

loki:/tmp# /usr/bin/ntpdate -bv hades
30 Apr 22:45:52 ntpdate[8833]: ntpdate 4.2.4p4 <at> 1.1520-o Fri Feb 22 18:37:11 UTC 2008 (1)
30 Apr 22:45:52 ntpdate[8833]: step time server 192.168.14.18 offset -5.728539 sec

so in effect yes :)

Hm. A five-second skip forwards via date --set works fine. A five-second
skip *backwards*, and instant boom.

(I suspect from earlier failures that sometimes time-slewing can cause
this too, because normally I run ntpd all the time, and that should only
slew the time under normal circumstances, never skip it: yet UML
2.6.24+ visited hang city every few days.)

I've been busy birthdaying today, but tomorrow I'll instrument things
and figure out just where a backward skip hangs and why (assuming you
haven't already done so: it's my own damn fault for being so laggardly,
I should have done this weeks ago).

(Continue reading)

WANG Cong | 2 May 08:36 2008
Picon

[Patch] (Resend) arch/um/kernel/irq.c: clean up some functions


Make activate_fd() and free_irq_by_irq_and_dev() static.
Remove init_aio_irq() since it has no users.

Compile testing passed.

Cc: Jeff Dike <jdike <at> addtoit.com>
Signed-off-by: WANG Cong <wangcong <at> zeuux.org>

---
  arch/um/include/irq_kern.h |    2 --
  arch/um/include/irq_user.h |    2 --
  arch/um/kernel/irq.c       |   35 ++---------------------------------
  3 files changed, 2 insertions(+), 37 deletions(-)

diff --git a/arch/um/include/irq_kern.h b/arch/um/include/irq_kern.h
index 4f77559..fba3895 100644
--- a/arch/um/include/irq_kern.h
+++ b/arch/um/include/irq_kern.h
 <at>  <at>  -13,8 +13,6  <at>  <at>  extern int um_request_irq(unsigned int irq, int fd, int type,
  			  irq_handler_t handler,
  			  unsigned long irqflags,  const char * devname,
  			  void *dev_id);
-extern int init_aio_irq(int irq, char *name,
-			irq_handler_t handler);

  #endif

diff --git a/arch/um/include/irq_user.h b/arch/um/include/irq_user.h
index e60b318..c6c784d 100644
(Continue reading)

WANG Cong | 2 May 08:39 2008
Picon

[Patch] (Resend) arch/um/kernel/mem.c: remove arch_validate()


Remove arch_validate(), because no one uses it.
Remove useless macro HAVE_ARCH_VALIDATE.
Make the variable 'empty_bad_page' static.

Compile testing passed.

Cc: Jeff Dike <jdike <at> addtoit.com>
Signed-off-by: WANG Cong <wangcong <at> zeuux.org>

---
  arch/um/kernel/ksyms.c |    1 -
  arch/um/kernel/mem.c   |   33 +--------------------------------
  include/asm-um/page.h  |    3 ---
  3 files changed, 1 insertions(+), 36 deletions(-)

diff --git a/arch/um/kernel/ksyms.c b/arch/um/kernel/ksyms.c
index 66e2a30..cc1b5f1 100644
--- a/arch/um/kernel/ksyms.c
+++ b/arch/um/kernel/ksyms.c
 <at>  <at>  -18,7 +18,6  <at>  <at>  EXPORT_SYMBOL(get_signals);
  EXPORT_SYMBOL(kernel_thread);
  EXPORT_SYMBOL(sys_waitpid);
  EXPORT_SYMBOL(flush_tlb_range);
-EXPORT_SYMBOL(arch_validate);

  EXPORT_SYMBOL(high_physmem);
  EXPORT_SYMBOL(empty_zero_page);
diff --git a/arch/um/kernel/mem.c b/arch/um/kernel/mem.c
index 2eea1ff..88fc1c8 100644
(Continue reading)

WANG Cong | 2 May 08:45 2008
Picon

[Patch] (Resend) UML: make several more things static


Make some variables and functions static, since they don't need
to be global.

Remove an unused function - arch/um/kernel/time.c::sched_clock().

Clean the style a bit as complained by checkpatch.pl.

Cc: Jeff Dike <jdike <at> addtoit.com>
Signed-off-by: WANG Cong <wangcong <at> zeuux.org>

---
  arch/um/include/skas/skas.h     |    1 -
  arch/um/include/um_uaccess.h    |    1 -
  arch/um/kernel/physmem.c        |    2 +-
  arch/um/kernel/ptrace.c         |    2 +-
  arch/um/kernel/time.c           |    8 --------
  arch/um/kernel/uaccess.c        |    2 +-
  arch/um/os-Linux/sigio.c        |    2 +-
  arch/um/os-Linux/signal.c       |    2 +-
  arch/um/os-Linux/skas/process.c |    2 +-
  arch/um/os-Linux/umid.c         |    2 +-
  arch/um/sys-i386/bugs.c         |    2 +-
  arch/um/sys-i386/checksum.S     |    5 ++---
  arch/um/sys-i386/ldt.c          |    4 ++--
  include/asm-um/ptrace-generic.h |    3 ---
  14 files changed, 12 insertions(+), 26 deletions(-)

diff --git a/arch/um/include/skas/skas.h b/arch/um/include/skas/skas.h
index b073f8a..64d2c74 100644
(Continue reading)

WANG Cong | 2 May 08:59 2008
Picon

[Patch] UML: Fix inconsistence due to tty_operation change


'put_char' of 'struct tty_operations' has changed from 'void' into 'int'.
This can also shut up comipler warnings.

Cc: Jeff Dike <jdike <at> addtoit.com>
Signed-off-by: WANG Cong <wangcong <at> zeuux.org>

---

diff --git a/arch/um/drivers/line.c b/arch/um/drivers/line.c
index 10b86e1..5047490 100644
--- a/arch/um/drivers/line.c
+++ b/arch/um/drivers/line.c
 <at>  <at>  -191,9 +191,9  <at>  <at>  void line_flush_chars(struct tty_struct *tty)
  	line_flush_buffer(tty);
  }

-void line_put_char(struct tty_struct *tty, unsigned char ch)
+int line_put_char(struct tty_struct *tty, unsigned char ch)
  {
-	line_write(tty, &ch, sizeof(ch));
+	return line_write(tty, &ch, sizeof(ch));
  }

  int line_write(struct tty_struct *tty, const unsigned char *buf, int len)
diff --git a/arch/um/include/line.h b/arch/um/include/line.h
index 1223f2c..979b73e 100644
--- a/arch/um/include/line.h
+++ b/arch/um/include/line.h
 <at>  <at>  -71,7 +71,7  <at>  <at>  extern int line_setup(struct line *lines, unsigned int sizeof_lines,
(Continue reading)


Gmane