Skip to contentSkip to conversation

2.6.16-rc2-mm1 patches don't apply

DISCUSSION Paul Jackson 4 participants 10 messages Feb 9, 2006 – Feb 18, 2006

lore
Share & tools
lore
#0 lore raw

Andrew,

The patchset for 2.6.16-rc2-mm1 seems borked.

It contains patches in linus.patch that are already in 2.6.16-rc2.

Hence, this *-mm patch set cannot be applied on top of 2.6.16-rc2.

For example, consider the following.

Linus tagged v2.6.16-rc2 at this revision (hg log format):

	changeset:   19933:6a79f5a2de38
	tag:         v2.6.16-rc2
	user:        Linus Torvalds <[email protected]>
	date:        Fri Feb  3 14:03:08 2006 +0800
	summary:     Linux v2.6.16-rc2

The following patch is included earlier in v2.6.16-rc2:

	changeset:   19738:81ca7cda8d38
	user:        Lucas Correia Villa Real <[email protected]>
	date:        Wed Feb  1 21:24:23 2006 +0000
	summary:     [ARM] 3284/1: S3C2400 - adds support to GPIO

It affects a few files, including the following change, obtained using the command hg diff -r 19737 -r 19738 arch/arm/mach-s3c2410/Makefile:

	diff -r 03e77bf674ae -r 81ca7cda8d38 arch/arm/mach-s3c2410/Makefile
	--- a/arch/arm/mach-s3c2410/Makefile    Wed Feb  1 21:07:28 2006 +0000
	+++ b/arch/arm/mach-s3c2410/Makefile    Wed Feb  1 21:24:23 2006 +0000
	@@ -10,9 +10,13 @@
	 obj-n                  :=
	 obj-                   :=

	+# S3C2400 support files
	+obj-$(CONFIG_CPU_S3C2400)  += s3c2400-gpio.o
	+
	 # S3C2410 support files

	 obj-$(CONFIG_CPU_S3C2410)  += s3c2410.o
	+obj-$(CONFIG_CPU_S3C2410)  += s3c2410-gpio.o
	 obj-$(CONFIG_S3C2410_DMA)  += dma.o

	 # Power Management support
	@@ -25,6 +29,7 @@
	 obj-$(CONFIG_CPU_S3C2440)  += s3c2440.o s3c2440-dsc.o
	 obj-$(CONFIG_CPU_S3C2440)  += s3c2440-irq.o
	 obj-$(CONFIG_CPU_S3C2440)  += s3c2440-clock.o
	+obj-$(CONFIG_CPU_S3C2440)  += s3c2410-gpio.o

	 # bast extras

The linus.patch in the broken out patch set:

	2.6/2.6.16-rc2/2.6.16-rc2-mm1/2.6.16-rc2-mm1-broken-out.tar.bz2

from below:

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/

contains the lines:

	diff --git a/arch/arm/mach-s3c2410/Makefile b/arch/arm/mach-s3c2410/Makefile
	index b4f1e05..1217bf0 100644
	--- a/arch/arm/mach-s3c2410/Makefile
	+++ b/arch/arm/mach-s3c2410/Makefile
	@@ -10,9 +10,13 @@ obj-m                        :=
	 obj-n                  :=
	 obj-                   :=

	+# S3C2400 support files
	+obj-$(CONFIG_CPU_S3C2400)  += s3c2400-gpio.o
	+
	 # S3C2410 support files

	 obj-$(CONFIG_CPU_S3C2410)  += s3c2410.o
	+obj-$(CONFIG_CPU_S3C2410)  += s3c2410-gpio.o
	 obj-$(CONFIG_S3C2410_DMA)  += dma.o

	 # Power Management support
	@@ -25,6 +29,7 @@ obj-$(CONFIG_PM_SIMTEC)          += pm-simtec.
	 obj-$(CONFIG_CPU_S3C2440)  += s3c2440.o s3c2440-dsc.o
	 obj-$(CONFIG_CPU_S3C2440)  += s3c2440-irq.o
	 obj-$(CONFIG_CPU_S3C2440)  += s3c2440-clock.o
	+obj-$(CONFIG_CPU_S3C2440)  += s3c2410-gpio.o

	 # bast extras

Looks familiar, huh?

That's the same change, in both linus.patch of 2.6.16-rc2-mm1, and in what Linus marked as 2.6.16-rc2.

I think that there are some more, perhaps dozens, of such duplicated patches. See my earlier lkml post Message-Id:
<[email protected]> in response to Andrew's 2.6.16-rc2-mm1 announcement, for circumstantial evidence of more such damage to this patch set.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#1 lore raw
Quoted 4 lines · Paul Jackson

The patchset for 2.6.16-rc2-mm1 seems borked.

It contains patches in linus.patch that are already in 2.6.16-rc2.

<wget, wget>

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc2/2.6.16-rc2-mm1/broken-out/linus.patch applies cleanly to ftp://ftp.kernel.org/pub/linux/kernel/v2.6/testing/linux-2.6.16-rc2.tar.bz2

#2 lore raw
Quoted 3 lines · Andrew

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc2/2.6.16-rc2-mm1/broken-out/linus.patch applies cleanly to ftp://ftp.kernel.org/pub/linux/kernel/v2.6/testing/linux-2.6.16-rc2.tar.bz2

Indeed. That combination does work.

And that's the same linus.patch as I extracted from:

~akpm/patches/2.6/2.6.16-rc2//2.6.16-rc2-mm1/2.6.16-rc2-mm1-broken-out.tar.bz2

So no problem with you linus.patch.

The difference -might- be a different linux-2.6.16-rc2 with the above than with hg's http://www.kernel.org/hg/linux-2.6, with the command "hg co 19933", which is the release tagged v2.6.16-rc2:

    changeset:   19933:6a79f5a2de38
    tag:         v2.6.16-rc2
    user:        Linus Torvalds <[email protected]>
    date:        Fri Feb  3 14:03:08 2006 +0800
    summary:     Linux v2.6.16-rc2

My current suspicions ... a combination of:

1) a congenital mental defect that causes me to find git difficult to use, and led to me thinking something had failed when it had not, and
2) further fallout from some apparent burp in the production of the hg (mercurial) linux trees that Matt fixed today.

I can't do much about (1). I will see if I can nail down (2), and send Matt details, if I get far enough to be useful.

Thanks, Andrew, for running your <wget, wget> test.

That got me back on the road again.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#3 lore raw
Quoted 23 lines · Paul Jackson
Quoted 3 lines · Andrew

ftp://ftp.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.16-rc2/2.6.16-rc2-mm1/broken-out/linus.patch applies cleanly to ftp://ftp.kernel.org/pub/linux/kernel/v2.6/testing/linux-2.6.16-rc2.tar.bz2

Indeed. That combination does work.

And that's the same linus.patch as I extracted from:

~akpm/patches/2.6/2.6.16-rc2//2.6.16-rc2-mm1/2.6.16-rc2-mm1-broken-out.tar.bz2

So no problem with you linus.patch.

The difference -might- be a different linux-2.6.16-rc2 with the above than with hg's http://www.kernel.org/hg/linux-2.6, with the command "hg co 19933", which is the release tagged v2.6.16-rc2:

    changeset:   19933:6a79f5a2de38
    tag:         v2.6.16-rc2
    user:        Linus Torvalds <[email protected]>
    date:        Fri Feb  3 14:03:08 2006 +0800
    summary:     Linux v2.6.16-rc2
Quoted 6 lines

My current suspicions ... a combination of:

1) a congenital mental defect that causes me to find git difficult to use, and led to me thinking something had failed when it had not, and
2) further fallout from some apparent burp in the production of the hg (mercurial) linux trees that Matt fixed today.

The git->hg conversion script may be borked right now. The upgrade of master.kernel.org to AMD64 happened to coincide with a new Mercurial release with a new convert-repo script which may have introduced a bug. I'll try to sort it out tomorrow.

Signature · 1 line
Mathematics is the supreme nostalgia of our time.
#4 lore raw
Quoted 1 line

The git->hg conversion script may be borked right now.

Ok - since you have suspicions, I will dump what I have found so far, and pass the baton to your good hands.

The following "facts" are not to be trusted too much, just what I think I have now.

The following 34 files seem to come up differently in the hg side than the git side, for 2.6.16-rc2, with the hg side having patches showing up as already in 2.6.16-rc2, that the git side things showed up later, after Linus tagged 2.6.16-rc2.

./arch/arm/mach-s3c2410/Makefile
./arch/arm/mm/proc-xscale.S
./arch/cris/Makefile
./arch/ia64/Kconfig
./arch/ia64/kernel/sal.c
./arch/ia64/sn/Makefile
./arch/ia64/sn/kernel/Makefile
./arch/ia64/sn/kernel/bte.c
./arch/ia64/sn/kernel/klconflib.c
./arch/ia64/sn/kernel/setup.c
./arch/ia64/sn/kernel/sn2/Makefile
./arch/ia64/sn/kernel/sn2/sn2_smp.c
./arch/ia64/sn/pci/Makefile
./arch/ia64/sn/pci/pcibr/Makefile
./arch/parisc/Kconfig.debug
./arch/parisc/hpux/entry_hpux.S
./arch/parisc/kernel/parisc_ksyms.c
./arch/parisc/kernel/perf_images.h
./arch/parisc/kernel/signal32.c
./arch/parisc/kernel/signal32.h
./arch/parisc/math-emu/decode_exc.c
./drivers/media/dvb/b2c2/Kconfig
./drivers/media/dvb/b2c2/flexcop-dma.c
./drivers/media/dvb/b2c2/flexcop-misc.c
./drivers/media/dvb/b2c2/flexcop-pci.c
./drivers/media/dvb/bt8xx/bt878.h
./drivers/net/wan/dscc4.c
./include/asm-arm/checksum.h
./include/asm-ia64/sal.h
./include/asm-ia64/sn/bte.h
./include/asm-ia64/system.h
./include/asm-parisc/compat_ucontext.h
./include/asm-parisc/pgalloc.h
./include/asm-parisc/rt_sigframe.h

The linus.patch of 2.6.16-rc2-mm1 won't apply to the git variant of 2.6.16-rc2, because that *-mm1 tries to apply the changes on the above files that are already present.

Good luck with it.

Thanks, Matt.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#5 lore raw

HG kernel repo (was: Re: 2.6.16-rc2-mm1 patches don't apply)

In reply to Matt Mackall

Hi Matt,

Quoted 4 lines

The git->hg conversion script may be borked right now. The upgrade of master.kernel.org to AMD64 happened to coincide with a new Mercurial release with a new convert-repo script which may have introduced a bug. I'll try to sort it out tomorrow.

Any news or ETA on this ?

Thanks,

Stelian.

Signature · 1 line
Stelian Pop <[email protected]>
#6 lore raw

Re: HG kernel repo (was: Re: 2.6.16-rc2-mm1 patches don't apply)

Quoted 8 lines · Stelian Pop

Hi Matt,

Quoted 4 lines

The git->hg conversion script may be borked right now. The upgrade of master.kernel.org to AMD64 happened to coincide with a new Mercurial release with a new convert-repo script which may have introduced a bug. I'll try to sort it out tomorrow.

Any news or ETA on this ?

Sorry folks, I've been in meetings all weekend.

Signature · 1 line
Mathematics is the supreme nostalgia of our time.
#7 lore raw

Re: HG kernel repo (was: Re: 2.6.16-rc2-mm1 patches don't apply)

Quoted 1 line

Sorry folks, I've been in meetings all weekend.

Any progress yet, Matt, on hg kernel repository?

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401
#8 lore raw

Re: HG kernel repo (was: Re: 2.6.16-rc2-mm1 patches don't apply)

Quoted 3 lines · Paul Jackson
Quoted 1 line

Sorry folks, I've been in meetings all weekend.

Any progress yet, Matt, on hg kernel repository?

Yeah, it's been fixed for several days now. Forgot to send an update.

Signature · 1 line
Mathematics is the supreme nostalgia of our time.
#9 lore raw

Re: HG kernel repo (was: Re: 2.6.16-rc2-mm1 patches don't apply)

Quoted 1 line

it's been fixed for several days now

good - thanks.

Signature · 3 lines
                  I won't rest till it's the best ...
                  Programmer, Linux Scalability
                  Paul Jackson <[email protected]> 1.925.600.0401