2013-09-16 22:08:14

by Linus Torvalds

[permalink] [raw]
Subject: Linux 3.12-rc1

So it's been two weeks, and the merge window for 3.12 is now closed.

The git trees have been updated, the tar-balls and patches should be
out too, and here's my "short mergelog" for the merge window: it's
kind of like "git shortlog", except it names the person I merged from
(_not_ necessarily the author of the actual work, but the maintainer
who sent me the pull request) along with a very short description of
what the pull was for.

In general, this merge window was fairly normal. About 73% drivers,
12% architecture updates, and 6% filesystems. The rest falls under
"misc".

I personally particularly like the scalability improvements that got
merged this time around. The tty layer locking got cleaned up and in
the process a lot of locking became per-tty, which actually shows up
on some (admittedly odd) loads. And the dentry refcount scalability
work means that the filename caches now scale very well indeed, even
for the case where you look up the same directory or file (which could
historically result in contention on the per-dentry d_lock).

But those things aren't noticeable on normal machines, I'm just odd
and tend to get excited about improvements to our dentry cache. Just
because it's one of the more interesting parts of the core code to me
personally.

So most other people will probably care more about all the driver
updates that actually affect more everyday life.

Go forth and test,

Linus

---

Al Viro (4):
vfs pile 1
vfs pile 2 (of many)
vfs pile 3 (of many)
vfs pile 4

Alex Williamson (1):
VFIO update

Andrew Morton (2):
first patch-bomb
more patches

Anton Vorontsov (1):
battery/power supply driver updates

Artem Bityutskiy (2):
ubifs fix
UBI fixes

Ben Herrenschmidt (2):
powerpc updates
powerpc fixes

Ben LaHaise (1):
aio changes

Ben Myers (2):
xfs updates
xfs update #2

Bjorn Helgaas (1):
PCI changes

Bruce Fields (1):
nfsd updates

Bryan Wu (1):
led updates

Catalin Marinas (1):
ARM64 update

Chris Ball (1):
MMC updates

Chris Mason (1):
btrfs updates

Chris Metcalf (1):
Tile arch updates

Chris Zankel (1):
Xtensa updates

Dan Williams (1):
dmaengine update

Dave Airlie (2):
drm tree changes
drm fixes

David Miller (5):
networking changes
sparc changes
IDE changes
networking fixes
networking fixes

David Teigland (1):
dlm updates

David Woodhouse (1):
mtd updates

Dmitry Torokhov (2):
input updates
input update

Eric Biederman (1):
namespace changes

Eric Van Hensbergen (1):
9p updates

Geert Uytterhoeven (1):
m68k updates

Gleb Natapov (1):
KVM updates

Grant Likely (1):
device tree core updates

Greg KH (5):
USB patches
char/misc patches
driver core patches
staging tree merge
tty/serial driver patches

Greg Ungerer (1):
m68knommu fixes

Guenter Roeck (3):
hwmon updates
hwmon cleanups
hwmon fixes

Heiko Carstens (1):
more s390 updates

Herbert Xu (2):
crypto update
crypto fixes

Ingo Molnar (23):
RCU updates
core/locking changes
perf changes
scheduler changes
timer changes
x86/apic changes
x86/asm changes
x86/asmlinkage changes
tiny x86 boot cleanups
x86 cpu feature fixes
x86 fb changes
timers/nohz changes
x86 relocation changes
x86 mm changes
x86 paravirt changes
x86 platform documentation fix
x86 RAS changes
x86 SMAP fixes
x86 spinlock changes
x86 fixes
cputime fix
scheduler fix
perf fixes

Jaegeuk Kim (1):
f2fs updates

James Bottomley (2):
first round of SCSI updates
misc SCSI driver updates

James Hogan (1):
metag architecture changes

James Morris (1):
security subsystem updates

Jan Kara (1):
ext3, reiserfs, udf & isofs fixes

Jean Delvare (1):
hwmon fixes

Jesper Nilsson (1):
CRIS updates

Jiri Kosina (2):
HID updates
trivial tree

Joerg Roedel (1):
IOMMU Updates

Jon Mason (1):
NTB (non-transparent bridge) updates

Kevin Hilman (3):
ARM SoC driver update
ARM Renesas SoC cleanup, refactoring and more SMP support
ARM SoC late changes

Konrad Rzeszutek Wilk (2):
Xen updates
Xen bug-fixes

Linus Walleij (2):
pin control changes
GPIO updates

Marek Szyprowski (2):
DMA mapping update
DMA-mapping fix

Mark Brown (3):
regmap updates
spi updates
regulator updates

Martin Schwidefsky (2):
first batch of s390 updates
generic hardirq option removal

Matthew Garrett (1):
x86 platform updates

Matthew Wilcox (1):
NVM Express driver update

Mauro Carvalho Chehab (1):
media updates

Michael Turquette (1):
clock framework changes

Michal Marek (4):
kbuild update
misc kbuild updates
kconfig updates
kconfig fix

Michal Simek (1):
Microblaze patches

Mike Snitzer (1):
device-mapper updates

Miklos Szeredi (1):
fuse bugfixes

Neil Brown (1):
md update

Nicholas Bellinger (2):
SCSI target fixes
SCSI target updates

Olof Johansson (6):
ARM SoC low-priority fixes
ARM SoC cleanups
ARM SoC DT updates
ARM SoC platform changes
ARM SoC board updates
ARM SoC fixes

Pekka Enberg (1):
SLAB update

Peter Anvin (1):
x86 jumplabel changes

Phillip Lougher (1):
squashfs updates

Rafael Wysocki (2):
ACPI and power management updates
ACPI and power management fixes

Ralf Baechle (2):
MIPS updates
MIPS fixes

Richard Weinberger (1):
UML updates

Roland Dreier (1):
main batch of InfiniBand/RDMA changes

Russell King (2):
ARM updates
ARM fixes

Rusty Russell (3):
PTR_RET() removal patches
module updates
virtio update

Sage Weil (1):
ceph updates

Samuel Ortiz (1):
MFD (multi-function device) updates

Stefan Richter (1):
firewire updates

Stefano Stabellini (1):
Xen balloon driver bug-fixes

Steve French (2):
CIFS fixes
CIFS fixes

Steven Miao (1):
blackfin updates

Steven Rostedt (1):
tracing updates

Steven Whitehouse (1):
GFS2 updates

Sumit Semwal (1):
dma-buf updates

Takashi Iwai (2):
sound updates
sound fixes

Ted Ts'o (1):
ext4 updates

Tejun Heo (4):
single percpu update
workqueue updates
libata changes
cgroup updates

Thierry Reding (1):
pwm changes

Thomas Gleixner (1):
timer code update

Tomi Valkeinen (2):
OMAP specific fbdev changes
fbdev changes

Tony Luck (2):
pstore changes
ia64 fixes

Trond Myklebust (2):
NFS client updates
NFS client bugfixes (part 2)

Tyler Hicks (1):
eCryptfs fixes

Vineet Gupta (1):
ARC changes

Vinod Koul (1):
slave-dmaengine updates

Wim Van Sebroeck (1):
watchdog updates

Wolfram Sang (1):
i2c updates

Wu Fengguang (1):
writeback fix

Zhang Rui (1):
thermal management updates


2013-09-17 05:51:13

by Stephen Rothwell

[permalink] [raw]
Subject: linux-next stats (Was: Linux 3.12-rc1)

On Mon, 16 Sep 2013 18:08:11 -0400 Linus Torvalds <[email protected]> wrote:
>
> So it's been two weeks, and the merge window for 3.12 is now closed.

As usual, the executive friendly graph is at
http://neuling.org/linux-next-size.html :-)

(No merge commits counted, next-20130903 was the linux-next based on v3.11)

Commits in v3.12-rc1 (relative to v3.11): 9474 (v3.11-rc11: 9494)
Commits in next-20130903: 8891 (next-20130701: 8929)
Commits with the same SHA1: 7991 ( 7670)
Commits with the same patch_id: 472 (1) ( 759)
Commits with the same subject line: 70 (1) ( 55)

(1) not counting those in the lines above.

So commits in -rc1 that were "in" next-20130903: 8533 90.1% (8484 89.4%)
Commits in -rc1 that were not in next-20120722: 941 9.9% (1010 10.6%

So better than last time, but it would be still nice to figure out where
the last lot came from. I have the "git log --oneline --no-walk" list if
someone wants them.

Some breakdown of that list:

Top ten first word of commit summary:

57 net
53 mips
49 drm
47 [scsi]
23 perf
23 nfs
20 cifs
19 nvme
18 vfs
17 arm

Top ten authors:

33 Al Viro <[email protected]>
21 Sachin Kamat <[email protected]>
20 Linus Torvalds <[email protected]>
18 James Smart <[email protected]>
17 Jon Mason <[email protected]>
17 Chuck Lever <[email protected]>
16 Tomasz Figa <[email protected]>
16 Jingoo Han <[email protected]>
15 Daniel Borkmann <[email protected]>
14 Trond Myklebust <[email protected]>

Top ten commiters:

162 David S. Miller <[email protected]>
64 Linus Torvalds <[email protected]>
53 Ralf Baechle <[email protected]>
53 Al Viro <[email protected]>
52 Trond Myklebust <[email protected]>
47 James Bottomley <[email protected]>
22 Steve French <[email protected]>
21 Inki Dae <[email protected]>
21 Arnaldo Carvalho de Melo <[email protected]>
19 Rafael J. Wysocki <[email protected]>

Quite a few of these could be bug fixes (especially DaveM's).

There are also 358 commits in next-20130701 that didn't make it into v3.11-rc1.

Top ten first word of commit summary:

56 arm
34 drm
23 selinux
15 drivers
13 ocfs2
9 iov_iter
9 bluetooth
7 kdb
6 pci
5 watchdog

Top ten authors:

31 Andrew Morton <[email protected]>
22 Dave Kleikamp <[email protected]>
15 Jani Nikula <[email protected]>
13 Sebastian Hesselbarth <[email protected]>
12 Eric Paris <[email protected]>
10 Joern Engel <[email protected]>
9 Zach Brown <[email protected]>
9 Paul Moore <[email protected]>
9 Marcel Holtmann <[email protected]>
8 Guenter Roeck <[email protected]>

Some of Andrew's patches are fixes for other patches in his tree (and
have been merged into those).

Top ten commiters:

116 Stephen Rothwell <[email protected]>
33 Dave Kleikamp <[email protected]>
31 Daniel Vetter <[email protected]>
28 Benoit Cousson <[email protected]>
26 Eric Paris <[email protected]>
22 Jason Cooper <[email protected]>
14 Joern Engel <[email protected]>
11 Shawn Guo <[email protected]>
10 Jason Wessel <[email protected]>
9 Gustavo Padovan <[email protected]>

Well, that's embarrassing again :-) Those commits by me are from the
quilt series (mainly Andrew's mmotm tree).

Some of the above will have been merged into other patches or replaced, I
guess.
--
Cheers,
Stephen Rothwell [email protected]


Attachments:
(No filename) (3.69 kB)
(No filename) (836.00 B)
Download all attachments

2013-09-17 07:26:47

by Sedat Dilek

[permalink] [raw]
Subject: Re: linux-next stats (Was: Linux 3.12-rc1)

Hi Stephen,

I am subscribed to linux-next but I did not receive [1].
If you did not CC linux-next, can you do this next time?
Thanks.

Your stats have a lot of number, I am not sure if I read/interpret
them correctly.
Just one number is interesting for me:
How many commits/patches were taken from Linux-next tree 1:1 into v3.12-rc1?
Percentage of them to the total number of commits v3.11..v3.12-rc1?
IOW: How successful is Linux-next?

I am irritated by the number of SHA-IDs vs. number of subject-lines.
If those patches went into v3.12-rc1 from Linux-next both should be
nearly the same.
Can you explain that?

How many trees of maintainers/submaintainers were taken from their
"for-next" trees [2]?
IOW: Without any changes, special preperations "for-linus" (aka
"for-v3.12-rc1")?
IOW #2: How effective work maintainers :-)!

To summarize: I am interested in what went from -next to -rc1 in the
last merge-window.

Thanks for clarification in advance.

Regards,
- Sedat -

P.S.: While digging through the scripts-dir I found "diffconfig". For
your stats I highly recommend to see which Kconfig-options are 1. new
2. changed (name or Y|M|N, especially M->Y).


[1] http://marc.info/?l=linux-kernel&m=137939708327120&w=2
[2] http://git.kernel.org/cgit/linux/kernel/git/next/linux-next.git/tree/Next/Trees

2013-09-18 00:16:03

by Nicholas A. Bellinger

[permalink] [raw]
Subject: Re: linux-next stats (Was: Linux 3.12-rc1)

Hi Stephen,

On Tue, 2013-09-17 at 15:50 +1000, Stephen Rothwell wrote:
> On Mon, 16 Sep 2013 18:08:11 -0400 Linus Torvalds <[email protected]> wrote:
> >
> > So it's been two weeks, and the merge window for 3.12 is now closed.
>
> As usual, the executive friendly graph is at
> http://neuling.org/linux-next-size.html :-)
>
> (No merge commits counted, next-20130903 was the linux-next based on v3.11)
>
> Commits in v3.12-rc1 (relative to v3.11): 9474 (v3.11-rc11: 9494)
> Commits in next-20130903: 8891 (next-20130701: 8929)
> Commits with the same SHA1: 7991 ( 7670)
> Commits with the same patch_id: 472 (1) ( 759)
> Commits with the same subject line: 70 (1) ( 55)
>
> (1) not counting those in the lines above.
>
> So commits in -rc1 that were "in" next-20130903: 8533 90.1% (8484 89.4%)
> Commits in -rc1 that were not in next-20120722: 941 9.9% (1010 10.6%
>
> So better than last time, but it would be still nice to figure out where
> the last lot came from. I have the "git log --oneline --no-walk" list if
> someone wants them.
>
> Some breakdown of that list:
>
> Top ten first word of commit summary:
>
> 57 net
> 53 mips
> 49 drm
> 47 [scsi]
> 23 perf
> 23 nfs
> 20 cifs
> 19 nvme
> 18 vfs
> 17 arm
>
> Top ten authors:
>
> 33 Al Viro <[email protected]>
> 21 Sachin Kamat <[email protected]>
> 20 Linus Torvalds <[email protected]>
> 18 James Smart <[email protected]>
> 17 Jon Mason <[email protected]>
> 17 Chuck Lever <[email protected]>
> 16 Tomasz Figa <[email protected]>
> 16 Jingoo Han <[email protected]>
> 15 Daniel Borkmann <[email protected]>
> 14 Trond Myklebust <[email protected]>
>

I'm totally confused by these stats..

The target-pending/for-next pull had ~30 commits with the term 'target'
in the first word of the commit summary, and yours truly had 40 commits
merged.

Is there a reason why these would not be showing up in the above..?

--nab

2013-09-18 00:38:05

by Stephen Rothwell

[permalink] [raw]
Subject: Re: linux-next stats (Was: Linux 3.12-rc1)

Hi Nicholas,

On Tue, 17 Sep 2013 17:23:41 -0700 "Nicholas A. Bellinger" <[email protected]> wrote:
>
> On Tue, 2013-09-17 at 15:50 +1000, Stephen Rothwell wrote:
> > On Mon, 16 Sep 2013 18:08:11 -0400 Linus Torvalds <[email protected]> wrote:
> > >
> > > So it's been two weeks, and the merge window for 3.12 is now closed.
> >
> > As usual, the executive friendly graph is at
> > http://neuling.org/linux-next-size.html :-)
> >
> > (No merge commits counted, next-20130903 was the linux-next based on v3.11)
> >
> > Commits in v3.12-rc1 (relative to v3.11): 9474 (v3.11-rc11: 9494)
> > Commits in next-20130903: 8891 (next-20130701: 8929)
> > Commits with the same SHA1: 7991 ( 7670)
> > Commits with the same patch_id: 472 (1) ( 759)
> > Commits with the same subject line: 70 (1) ( 55)
> >
> > (1) not counting those in the lines above.
> >
> > So commits in -rc1 that were "in" next-20130903: 8533 90.1% (8484 89.4%)
> > Commits in -rc1 that were not in next-20120722: 941 9.9% (1010 10.6%
> >
> > So better than last time, but it would be still nice to figure out where
> > the last lot came from. I have the "git log --oneline --no-walk" list if
> > someone wants them.
> >
> > Some breakdown of that list:
> >
> > Top ten first word of commit summary:
> >
> > 57 net
> > 53 mips
> > 49 drm
> > 47 [scsi]
> > 23 perf
> > 23 nfs
> > 20 cifs
> > 19 nvme
> > 18 vfs
> > 17 arm
> >
> > Top ten authors:
> >
> > 33 Al Viro <[email protected]>
> > 21 Sachin Kamat <[email protected]>
> > 20 Linus Torvalds <[email protected]>
> > 18 James Smart <[email protected]>
> > 17 Jon Mason <[email protected]>
> > 17 Chuck Lever <[email protected]>
> > 16 Tomasz Figa <[email protected]>
> > 16 Jingoo Han <[email protected]>
> > 15 Daniel Borkmann <[email protected]>
> > 14 Trond Myklebust <[email protected]>
>
> I'm totally confused by these stats..
>
> The target-pending/for-next pull had ~30 commits with the term 'target'
> in the first word of the commit summary, and yours truly had 40 commits
> merged.
>
> Is there a reason why these would not be showing up in the above..?

These are the lists if things that went into Linus' tree but were *not*
in linux-next prior to the merge window opening.

--
Cheers,
Stephen Rothwell [email protected]


Attachments:
(No filename) (2.42 kB)
(No filename) (836.00 B)
Download all attachments

2013-09-18 06:32:54

by Geert Uytterhoeven

[permalink] [raw]
Subject: Re: linux-next stats (Was: Linux 3.12-rc1)

On Wed, Sep 18, 2013 at 2:37 AM, Stephen Rothwell <[email protected]> wrote:
>> > So better than last time, but it would be still nice to figure out where
>> > the last lot came from. I have the "git log --oneline --no-walk" list if
>> > someone wants them.
>> >
>> > Some breakdown of that list:

>> I'm totally confused by these stats..
>>
>> The target-pending/for-next pull had ~30 commits with the term 'target'
>> in the first word of the commit summary, and yours truly had 40 commits
>> merged.
>>
>> Is there a reason why these would not be showing up in the above..?
>
> These are the lists if things that went into Linus' tree but were *not*
> in linux-next prior to the merge window opening.

Perhaps it should be made a little bit clearer that it's a hall of shame, not a
hall of fame? If you miss "the last lot", and continue reading to the various
"top ten"s, you really think you want to be part of them...

The first time they got published, I was also caught by this :-)

Gr{oetje,eeting}s,

Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds

2013-09-18 12:53:50

by Jason Cooper

[permalink] [raw]
Subject: Re: linux-next stats (Was: Linux 3.12-rc1)

Stephen,

On Tue, Sep 17, 2013 at 03:50:57PM +1000, Stephen Rothwell wrote:
> There are also 358 commits in next-20130701 that didn't make it into v3.11-rc1.
>
> Top ten first word of commit summary:
>
> 56 arm
>
> Top ten authors:
>
> 13 Sebastian Hesselbarth <[email protected]>
>
> Top ten commiters:
>
> 22 Jason Cooper <[email protected]>

:( That's my mistake. I should've been tracking my pull requests to
arm-soc more closely. We've already tweaked the process to prevent this
in the future.

thx,

Jason.

2013-09-19 15:45:35

by Mark Brown

[permalink] [raw]
Subject: Re: linux-next stats (Was: Linux 3.12-rc1)

On Tue, Sep 17, 2013 at 03:50:57PM +1000, Stephen Rothwell wrote:

> 21 Sachin Kamat <[email protected]>

These should all be fixes, Sachin submitted a bunch of fixes for common
patterns with errors just about the time the merge window opened.


Attachments:
(No filename) (255.00 B)
signature.asc (836.00 B)
Digital signature
Download all attachments

2013-09-23 14:34:01

by Rob Landley

[permalink] [raw]
Subject: Re: Linux 3.12-rc1

On 09/16/2013 05:08:11 PM, Linus Torvalds wrote:
> So it's been two weeks, and the merge window for 3.12 is now closed.
>
> The git trees have been updated, the tar-balls and patches should be
> out too, and here's my "short mergelog" for the merge window: it's
> kind of like "git shortlog", except it names the person I merged from
> (_not_ necessarily the author of the actual work, but the maintainer
> who sent me the pull request) along with a very short description of
> what the pull was for.

So this log is basically:

git log v3.11..v3.12-rc1 --merges --author="Linus Torvalds" \
| sed -n 's/ *Pull \(.*\) from \(.*\):/\2: \1/p' | sort

Only with a slightly more clever "sort"?

Rob-

2013-09-23 16:36:55

by Linus Torvalds

[permalink] [raw]
Subject: Re: Linux 3.12-rc1

On Mon, Sep 23, 2013 at 7:33 AM, Rob Landley <[email protected]> wrote:
>
> So this log is basically:
>
> git log v3.11..v3.12-rc1 --merges --author="Linus Torvalds" \
> | sed -n 's/ *Pull \(.*\) from \(.*\):/\2: \1/p' | sort
>
> Only with a slightly more clever "sort"?

Yes, except it's written in perl to get that simpler sort.

My silly script is designed to mimic "git shortlog" output, and it is
indeed based on the original perl version (and syntax) of that.

It also matches "Merge" in addition to "Pull", since some of my merges
end up being branches that get created from emails in my local
repository and then merged.

In case anybody wants to play with it, the perl script is attached,
and you can see the git-shortlog roots (for example, it still has the
usage string that mentions the -[hns] flags that don't actually work,
and some function called shortlog).

NOTE! It only works with the syntax that I use for merge messages, so
while you could use it to summarize other peoples pulls and merges, it
really doesn't end up working very well for that. Maybe some
developers have adopted my syntax, but a quick look says no. Anyway,
for my use it's

git log v3.11.. --merges --author=Torvalds | git-mergelog

and you can try to play around with other merge authors and maybe
accept a wider range of syntax, but from a quick look it looks like
none of the other maintainers that do merges have a very convenient
fixed format..

And my perl-fu is lacking, so don't laugh at my script. But if you
have improvements, feel free to send them...

Linus


Attachments:
git-mergelog (1.67 kB)