Very Computer

        Randy Dunlap
        Andi Kleen
        Christoph Hellwig
        Kai Germaschewski
        Stephen Hemminger

We have tested again on multiple kernels with various types of
dumping devices.  The diffstat output is now at the top of each
patch.  These patches have been well tested.  Feel free to give
them a spin and let us know of any problems.

Thanks,

--Matt

The full diffstat for LKCD in 2.5.44 tree (the majority of the
changes are the addition of the dump modules and headers):

 Makefile                             |   15
 arch/i386/boot/Makefile              |    2
 arch/i386/boot/install.sh            |   24
 arch/i386/config.in                  |    7
 arch/i386/kernel/Makefile            |    2
 arch/i386/kernel/irq.c               |    5
 arch/i386/kernel/nmi.c               |    9
 arch/i386/kernel/smp.c               |   15
 arch/i386/kernel/traps.c             |   28
 arch/i386/mach-generic/irq_vectors.h |    1
 arch/i386/mm/Makefile                |    2
 arch/i386/mm/init.c                  |    5
 arch/s390/boot/install.sh            |   24
 arch/s390x/boot/install.sh           |   24
 drivers/Makefile                     |    1
 drivers/char/sysrq.c                 |   13
 drivers/dump/Makefile                |   30
 drivers/dump/dump_base.c             | 1860 +++++++++++++++++++++++++++++++++++
 drivers/dump/dump_blockdev.c         |  392 +++++++
 drivers/dump/dump_gzip.c             |  129 ++
 drivers/dump/dump_i386.c             |  315 +++++
 drivers/dump/dump_rle.c              |  176 +++
 include/asm-i386/dump.h              |   94 +
 include/asm-i386/smp.h               |    1
 include/linux/dump.h                 |  438 ++++++++
 include/linux/page-flags.h           |    5
 init/Makefile                        |    5
 init/kerntypes.c                     |   24
 init/main.c                          |   10
 kernel/Makefile                      |    2
 kernel/panic.c                       |   16
 kernel/sched.c                       |   30
 kernel/sys.c                         |   36
 lib/Config.in                        |    2
 mm/page_alloc.c                      |   22
 35 files changed, 3731 insertions(+), 33 deletions(-)
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

                        NOT

--Matt

|>> These are the latest LKCD patches for 2.5.44.  We have put in a
|>> number of changes, additions and deletions requested by the
|>> following people on the lkml list between our 2.5.38 and 2.5.44
|>> patches:
|>
|>Thank you for being so polite to actually make your patches into 1
|>thread only as I asked yesterday and a lot of people agreed to.
|>
|>                   NOT

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

I have no real interest in LKCD functionality. I'm really only concerned that you aren't making my life ugly.

And you are, in a couple of small but annoying ways:
1. No Config.help entry. So I still have no idea what LKCD does, and whether I need to enable it in my kernel.

2. Needless compile time errors.
make -f drivers/dump/Makefile
  gcc -Wp,-MD,drivers/dump/.dump_base.o.d -D__KERNEL__ -Iinclude -Wall -Wstrict-prototypes -Wno-trigra
phs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -marc
h=i686 -Iarch/i386/mach-generic -nostdinc -iwithprefix include    -DKBUILD_BASENAME=dump_base -DEXPORT
_SYMTAB  -c -o drivers/dump/dump_base.o drivers/dump/dump_base.c
In file included from drivers/dump/dump_base.c:204:
include/linux/version.h:3: warning: `KERNEL_VERSION' redefined
include/linux/dump.h:434: warning: this is the location of the previous definition
  gcc -Wp,-MD,drivers/dump/.dump_i386.o.d -D__KERNEL__ -Iinclude -Wall -Wstrict-prototypes -Wno-trigra
phs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -marc
h=i686 -Iarch/i386/mach-generic -nostdinc -iwithprefix include    -DKBUILD_BASENAME=dump_i386   -c -o
drivers/dump/dump_i386.o drivers/dump/dump_i386.c
  ld -m elf_i386  -r -o drivers/dump/dump.o drivers/dump/dump_base.o drivers/dump/dump_i386.o
  gcc -Wp,-MD,drivers/dump/.dump_blockdev.o.d -D__KERNEL__ -Iinclude -Wall -Wstrict-prototypes -Wno-tr
igraphs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -
march=i686 -Iarch/i386/mach-generic -nostdinc -iwithprefix include    -DKBUILD_BASENAME=dump_blockdev
  -c -o drivers/dump/dump_blockdev.o drivers/dump/dump_blockdev.c
drivers/dump/dump_blockdev.c:14: warning: `DUMP_MODULE_NAME' redefined
include/linux/dump.h:403: warning: this is the location of the previous definition
  gcc -Wp,-MD,drivers/dump/.dump_rle.o.d -D__KERNEL__ -Iinclude -Wall -Wstrict-prototypes -Wno-trigrap
hs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -march
=i686 -Iarch/i386/mach-generic -nostdinc -iwithprefix include    -DKBUILD_BASENAME=dump_rle   -c -o dr
ivers/dump/dump_rle.o drivers/dump/dump_rle.c
  gcc -Wp,-MD,drivers/dump/.dump_gzip.o.d -D__KERNEL__ -Iinclude -Wall -Wstrict-prototypes -Wno-trigra
phs -O2 -fomit-frame-pointer -fno-strict-aliasing -fno-common -pipe -mpreferred-stack-boundary=2 -marc
h=i686 -Iarch/i386/mach-generic -nostdinc -iwithprefix include    -DKBUILD_BASENAME=dump_gzip   -c -o
drivers/dump/dump_gzip.o drivers/dump/dump_gzip.c
drivers/dump/dump_gzip.c:27: warning: `DUMP_MODULE_NAME' redefined
include/linux/dump.h:403: warning: this is the location of the previous definition
drivers/dump/dump_gzip.c:29: warning: `DUMP_PRINTN' redefined
include/linux/dump.h:413: warning: this is the location of the previous definition
drivers/dump/dump_gzip.c:31: warning: `DUMP_PRINT' redefined
include/linux/dump.h:431: warning: this is the location of the previous definition

And MOST IMPORTANTLY:
3. Kernel gets bigger, even with the config option turned off:
Baseline:
- -rwxr-xr-x    1 bradh    users     4542815 Oct 21 21:58 vmlinux
- -rw-r--r--    1 bradh    users     1571320 Oct 21 21:58 bzImage

With LKCD patches installed, and the same .config (except for LKCD stuff), it looks like:
- -rwxr-xr-x    1 bradh    users     4548357 Oct 21 22:19 vmlinux-configM
- -rwxr-xr-x    1 bradh    users     4548357 Oct 21 22:03 vmlinux-configN
- -rwxr-xr-x    1 bradh    users     4588663 Oct 21 22:21 vmlinux-configY
- -rw-r--r--    1 bradh    users     1571988 Oct 21 22:17 bzImage-configM
- -rw-r--r--    1 bradh    users     1571993 Oct 21 22:03 bzImage-configN
- -rw-r--r--    1 bradh    users     1590318 Oct 21 22:21 bzImage-configY

Where the -configX shows the effect of compiling with any options presented set to X (where X is N, M or Y).

I don't know enough VM or scheduler to critique the impact on performance with it turned on. However the very least you can do is to have no impact when your code is turned off.

Brad

- --
http://linux.conf.au. 22-25Jan2003. Perth, Aust. I'm registered. Are you?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.6 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE9s/RpW6pHgIdAuOMRAgodAJ9J/Lm2ShxMO/K7EkFRghkT9MF3gQCdGBpf
6d7H2aXsPgHsqnVOZwwJDGg=
=Ad25
-----END PGP SIGNATURE-----

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/