* 2.6.35-rc3: Reported regressions from 2.6.34
From: Rafael J. Wysocki @ 2010-06-20 22:11 UTC (permalink / raw)
To: Linux Kernel Mailing List
Cc: Maciej Rutecki, Andrew Morton, Linus Torvalds,
Kernel Testers List, Network Development, Linux ACPI,
Linux PM List, Linux SCSI List, Linux Wireless List, DRI
This message contains a list of some regressions from 2.6.34,
for which there are no fixes in the mainline known to the tracking team.
If any of them have been fixed already, please let us know.
If you know of any other unresolved regressions from 2.6.34, please let us
know either and we'll add them to the list. Also, please let us know
if any of the entries below are invalid.
Each entry from the list will be sent additionally in an automatic reply
to this message with CCs to the people involved in reporting and handling
the issue.
Listed regressions statistics:
Date Total Pending Unresolved
----------------------------------------
2010-06-21 46 37 26
2010-06-09 15 13 10
Unresolved regressions
----------------------
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16255
Subject : 2.6.35-rc3 deadlocks on semaphore operations
Submitter : Christoph Lameter <cl-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b@public.gmane.org>
Date : 2010-06-18 14:49 (3 days old)
Message-ID : <alpine.DEB.2.00.1006180940140.11575-sBS69tsa9Uj/9pzu0YdTqQ@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127687262727707&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16248
Subject : inconsistent lock state
Submitter : Sergey Senozhatsky <sergey.senozhatsky-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-15 11:24 (6 days old)
Message-ID : <20100615112434.GA3967-dY8u8AhHFaWtd10JCjopabkcH5ONE+aC@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127660087625903&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16247
Subject : drm/i915 BUG with 2.6.35-rc
Submitter : Benny Halevy <bhalevy-C4P08NqkoRlBDgjK7y7TUQ@public.gmane.org>
Date : 2010-06-14 22:38 (7 days old)
Message-ID : <4C16AF56.1040002-C4P08NqkoRlBDgjK7y7TUQ@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127655510531367&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16235
Subject : [REGRESSION] [IWL3945] Broadcast is broken?
Submitter : Maciej Rutecki <maciej.rutecki-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-14 17:24 (7 days old)
Message-ID : <201006141924.24061.maciej.rutecki-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127653628301300&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16234
Subject : [2.6.35-rc3] reboot mutex 'bug'...
Submitter : Daniel J Blueman <daniel.blueman-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-14 15:16 (7 days old)
Message-ID : <AANLkTimDcTnyEPmt2ZcCM1UWtn4AYKotiqyjobJApkO7-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127652861118933&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16232
Subject : 2.6.35-rc3 - WARNING:iwl_set_dynamic_key
Submitter : Mario Guenterberg <mario.guenterberg-gM/Ye1E23mwN+BqQ9rBEUg@public.gmane.org>
Date : 2010-06-14 11:55 (7 days old)
Message-ID : <20100614115510.GA7904@guenti-laptop>
References : http://marc.info/?l=linux-kernel&m=127651695627147&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16230
Subject : inconsistent IN-HARDIRQ-W -> HARDIRQ-ON-W usage: fasync, 2.6.35-rc3
Submitter : Dominik Brodowski <linux-X3ehHDuj6sIIGcDfoQAp7OTW4wlIGRCZ@public.gmane.org>
Date : 2010-06-13 9:53 (8 days old)
Message-ID : <20100613095305.GA13231-S7uyTPAaJ/sb6pqDj42GsMgv3T4z79SOrE5yTffgRl4@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127642282208277&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16228
Subject : BUG/boot failure on Dell Precision T3500 (pci/ahci_stop_engine)
Submitter : Brian Bloniarz <phunge0-PkbjNfxxIARBDgjK7y7TUQ@public.gmane.org>
Date : 2010-06-16 17:57 (5 days old)
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16222
Subject : 2.6.35-rc{12} regression: inactive console corrupted
Submitter : Pavel Machek <pavel-+ZI9xUNit7I@public.gmane.org>
Date : 2010-06-12 10:33 (9 days old)
Message-ID : <20100612103321.GA1458-+ZI9xUNit7I@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127633882614501&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16221
Subject : 2.6.35-rc2-git5 -- [drm:drm_mode_getfb] *ERROR* invalid framebuffer id
Submitter : Miles Lane <miles.lane-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-11 20:31 (10 days old)
Message-ID : <AANLkTim0jVRyqkwlGOcrg_XTvUQwcBYfWJX-aRzkkrLG-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127628828119623&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16205
Subject : acpi: freeing invalid memtype bf799000-bf79a000
Submitter : Marcin Slusarz <marcin.slusarz-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-09 20:09 (12 days old)
Message-ID : <20100609200910.GA2876-OI9uyE9O0yo@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127611427029914&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16201
Subject : SIOCGIWFREQ ioctl fails to get frequency info
Submitter : nuh <nuh-hRtevi7K+EWJQ7yn63+t2w@public.gmane.org>
Date : 2010-06-14 10:45 (7 days old)
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16199
Subject : 2.6.35-rc2-git1 - include/linux/cgroup.h:534 invoked rcu_dereference_check() without protection!
Submitter : Miles Lane <miles.lane-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-07 18:14 (14 days old)
Message-ID : <AANLkTin2pPqOUx--9fIX3BH3e-cU6oCRufijcx_4ozx5-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127593447812015&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16198
Subject : Running make install over sshfs is painful now
Submitter : Dmitry Torokhov <dmitry.torokhov-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-07 6:53 (14 days old)
Message-ID : <20100607065324.GA25590-WlK9ik9hQGAhIp7JRqBPierSzoNAToWh@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127589362011138&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16197
Subject : [BUG on 2.6.35-rc2] sysfs: cannot create duplicate filename '/devices/pci0000:00/0000:00:11.0/0000:02:03.0/slot'
Submitter : Ryan Wang <openspace.wang-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-07 0:23 (14 days old)
Message-ID : <AANLkTincwMZPnYW3S4uz4k2GOn52RpgBIBRfzyD010Yo-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127587022219378&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16184
Subject : Container, X86-64, i386, iptables rule
Submitter : Jean-Marc Pigeon <jmp-4qkeo2rQ0gg@public.gmane.org>
Date : 2010-06-12 04:17 (9 days old)
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16179
Subject : 2.6.35-rc2 completely hosed on intel gfx?
Submitter : Norbert Preining <preining-DX+603jRYB8@public.gmane.org>
Date : 2010-06-06 11:55 (15 days old)
Message-ID : <20100606115534.GA9399-DqSSrKF0TaySnEC3TeqHn5dqbFPxfnh/@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127582534931581&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16178
Subject : 2.6.35-rc2 : OOPS with LTP memcg regression test run.
Submitter : Sachin Sant <sachinp-xthvdsQ13ZrQT0dZR+AlfA@public.gmane.org>
Date : 2010-06-06 15:18 (15 days old)
Message-ID : <4C0BB98E.9030101-xthvdsQ13ZrQT0dZR+AlfA@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127583683912607&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16176
Subject : Microcode errors with iwl3945
Submitter : Steinar H. Gunderson <sgunderson-jG/AHqQBv7lBDgjK7y7TUQ@public.gmane.org>
Date : 2010-06-10 19:28 (11 days old)
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16175
Subject : 2.6.35-rc1 system oom, many processes killed but memory not free
Submitter : andrew hendry <andrew.hendry-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-05 0:46 (16 days old)
Message-ID : <AANLkTim7CiW-yfugZUAHZCqLvXKgt9CwolCvbLGdCLAk-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127569877714937&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16173
Subject : After uncompressing the kernel, at boot time, the server hangs.
Submitter : David Hill <hilld-HTiBYHdybX7UkGsOFmftXw@public.gmane.org>
Date : 2010-06-09 23:25 (12 days old)
First-Bad-Commit: http://git.kernel.org/linus/cf7500c0ea133d66f8449d86392d83f840102632
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16172
Subject : 2.6.25-rc1 ahci regression
Submitter : Luming Yu <luming.yu-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-03 7:02 (18 days old)
First-Bad-Commit: http://git.kernel.org/linus/365cfa1ed5a36f9bcb9f64c9f0f52155af2e9fef
Message-ID : <AANLkTikWak8i-Kaj-e-EUFWE66tfMmp2UPhGJCKkK1zK-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>>
References : http://marc.info/?l=linux-kernel&m=127554855724361&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16160
Subject : 2.6.35 Radeon KMS power management regression?
Submitter : Nigel Cunningham <ncunningham-VWkGxi2CTS+6c6uEtOJ/EA@public.gmane.org>
Date : 2010-06-01 6:23 (20 days old)
Message-ID : <4C04A767.8000209-VWkGxi2CTS+6c6uEtOJ/EA@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127537343722290&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16145
Subject : Unable to boot unless "notsc" or "clocksource=hpet", or acpi_pad disabling the TSC
Submitter : Tom Gundersen <teg-B22kvLQNl6c@public.gmane.org>
Date : 2010-06-07 13:11 (14 days old)
Handled-By : Venkatesh Pallipadi <venki-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
Len Brown <lenb-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16122
Subject : 2.6.35-rc1: WARNING at fs/fs-writeback.c:1142 __mark_inode_dirty+0x103/0x170
Submitter : Larry Finger <Larry.Finger-tQ5ms3gMjBLk1uMJSBkQmQ@public.gmane.org>
Date : 2010-06-04 13:18 (17 days old)
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16090
Subject : sysfs: cannot create duplicate filename
Submitter : Tobias <devnull-fBT1nhYaLZ4@public.gmane.org>
Date : 2010-06-01 15:59 (20 days old)
Handled-By : Jesse Barnes <jbarnes-Y1mF5jBUw70BENJcbMCuUQ@public.gmane.org>
Regressions with patches
------------------------
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16231
Subject : Noticeable slow-down in 2.6.35-rc3
Submitter : Chris Clayton <chris2553-gM/Ye1E23mwN+BqQ9rBEUg@public.gmane.org>
Date : 2010-06-13 20:15 (8 days old)
First-Bad-Commit: http://git.kernel.org/linus/597a264b1a9c7e36d1728f677c66c5c1f7e3b837
Message-ID : <AANLkTimDUf0N6b5PBy2wTGeAWEIY-lUvQPHQ7D7Gyrq4-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127646015226601&w=2
Handled-By : John Fastabend <john.r.fastabend-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Patch : https://patchwork.kernel.org/patch/105859/
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16220
Subject : md/raid/udev fails to create md partition devices (/dev/mdXpY)
Submitter : Duncan <1i5t5.duncan-j9pdmedNgrk@public.gmane.org>
Date : 2010-06-15 17:19 (6 days old)
Handled-By : Neil Brown <neilb-l3A5Bk7waGM@public.gmane.org>
Patch : https://bugzilla.kernel.org/show_bug.cgi?id=16220#c2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16215
Subject : sysfs: cannot create duplicate filename '/class/net/bnep0'
Submitter : Janusz Krzysztofik <jkrzyszt-NCk8gXQAEuFz6jiHbVrK7g@public.gmane.org>
Date : 2010-06-15 14:55 (6 days old)
Handled-By : Eric W. Biederman <ebiederm-aS9lmoZGLiVWk0Htik3J/w@public.gmane.org>
Patch : https://bugzilla.kernel.org/show_bug.cgi?id=16215#c10
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16196
Subject : 2.6.35-rc2-git1 - WARNING: at net/wireless/mlme.c:341 cfg80211_send_assoc_timeout+0x107/0x11e [cfg80211]()
Submitter : Miles Lane <miles.lane-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-07 19:06 (14 days old)
Message-ID : <AANLkTinmZh9GnrZLzOGi5Q_xHbky8cKdHAkOnsIA6_dZ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://marc.info/?l=linux-kernel&m=127593758817428&w=2
Handled-By : Johannes Berg <johannes.berg-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Patch : http://marc.info/?l=linux-wireless&m=127608099427316&w=2
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16180
Subject : radeon regression with 2.6.35-rc2-00001-g386f40c: black screen after resume
Submitter : cedric <cedric-x1Cn44Nr1HaZIoH1IeqzKA@public.gmane.org>
Date : 2010-06-11 10:59 (10 days old)
Handled-By : Cedric Godin <cedric.godin-AgBVmzD5pcezQB+pC5nmwQ@public.gmane.org>
Patch : https://bugzilla.kernel.org/attachment.cgi?id=26733
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16169
Subject : Complain from preemptive debug
Submitter : Sedat Dilek <sedat.dilek-gM/Ye1E23mwN+BqQ9rBEUg@public.gmane.org>
Date : 2010-05-31 10:10 (21 days old)
Message-ID : <AANLkTilTnAAZIizKinYsxFkNTkrmPqk6XJJuUjjhJ7EP-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
References : http://lkml.org/lkml/2010/5/31/77
Handled-By : Dmitry Monakhov <dmonakhov-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>
Patch : http://www.spinics.net/lists/cpufreq/msg01631.html
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16156
Subject : No Sound with 2.6.34.x Kernel with my Audio-Chip
Submitter : Ch: Hanisch <ch-hanisch-zqRNUXuvxA0b1SvskN2V4Q@public.gmane.org>
Date : 2010-06-08 11:51 (13 days old)
Handled-By : Takashi Iwai <tiwai-l3A5Bk7waGM@public.gmane.org>
Patch : https://bugzilla.kernel.org/attachment.cgi?id=26724
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16131
Subject : kernel BUG at fs/btrfs/extent-tree.c:4363 (btrfs_free_tree_block)
Submitter : Chow Loong Jin <hyperair-GeWIH/nMZzLQT0dZR+AlfA@public.gmane.org>
Date : 2010-06-05 18:53 (16 days old)
Handled-By : Yan Zheng <zheng.yan-QHcLZuEGTsvQT0dZR+AlfA@public.gmane.org>
Patch : https://patchwork.kernel.org/patch/103235/
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16127
Subject : Boot freeze on HP Compaq nx6325 (RS482) with Radeon KMS
Submitter : Jure Repinc <jlp.bugs-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-04 21:14 (17 days old)
Handled-By : Dave Airlie <airlied-cv59FeDIM0c@public.gmane.org>
Patch : https://bugzilla.kernel.org/attachment.cgi?id=26677
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16120
Subject : Oops: 0000 [#1] SMP, unable to handle kernel NULL pointer dereference at (null)
Submitter : Alex Zhavnerchik <alex.vizor-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Date : 2010-06-04 09:25 (17 days old)
Handled-By : Eric Dumazet <eric.dumazet-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
Patch : https://bugzilla.kernel.org/show_bug.cgi?id=16120#c6
Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=16092
Subject : Caught 64-bit read from uninitialized memory in memtype_rb_augment_cb
Submitter : Christian Casteyde <casteyde.christian-GANU6spQydw@public.gmane.org>
Date : 2010-06-01 18:08 (20 days old)
Handled-By : Venki <venki-hpIqsD4AKlfQT0dZR+AlfA@public.gmane.org>
Patch : https://bugzilla.kernel.org/show_bug.cgi?id=16092#c2
For details, please visit the bug entries and follow the links given in
references.
As you can see, there is a Bugzilla entry for each of the listed regressions.
There also is a Bugzilla entry used for tracking the regressions from 2.6.34,
unresolved as well as resolved, at:
http://bugzilla.kernel.org/show_bug.cgi?id=16055
Please let the tracking team know if there are any Bugzilla entries that
should be added to the list in there.
Thanks!
^ permalink raw reply
* Re: [PATCH] net: optimize Berkeley Packet Filter (BPF) processing
From: David Miller @ 2010-06-20 22:09 UTC (permalink / raw)
To: hagen; +Cc: shemminger, netdev
In-Reply-To: <20100620095019.GA32223@nuttenaction>
From: Hagen Paul Pfeifer <hagen@jauu.net>
Date: Sun, 20 Jun 2010 11:50:20 +0200
> * Stephen Hemminger | 2010-06-19 22:16:11 [-0700]:
>
>>I don't think this works because it breaks ABI compatibility for applications tha
>>use older versions.
>
> Are you sure Stephen? It is a one-to-one mapping of the ABI but maybe it was
> too late yesterday ... ;-)
I think from this perspective, the change is fine too, nothing visible
to the user is being changed here.
^ permalink raw reply
* Re: [net-next-2.6 PATCH 1/8] e1000e: cleanup ethtool loopback setup code
From: David Miller @ 2010-06-20 21:48 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, bphilips, bruce.w.allan
In-Reply-To: <AANLkTik2Rait7ugcDt36ncwqdO-fZrr9iM8usecAyBAC@mail.gmail.com>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Sun, 20 Jun 2010 00:32:20 -0700
> I know that Bruce (and company) are actively looking into your first
> point. I would think that he would have either a response or update
> on the status come Monday/Tuesday.
>
> Regarding option 2, understood and I will work out with Bruce on when
> and who will do the work. I will be taking a look at it tomorrow
> (later today), if I am not able to come up with a patch in the near
> future. I will return to working on it by Wednesday.
Great, thanks for letting me know it's being looked at.
^ permalink raw reply
* Re: [PATCH] net: optimize Berkeley Packet Filter (BPF) processing
From: Stephen Hemminger @ 2010-06-20 18:49 UTC (permalink / raw)
To: Hagen Paul Pfeifer; +Cc: netdev, davem
In-Reply-To: <20100620095019.GA32223@nuttenaction>
On Sun, 20 Jun 2010 11:50:20 +0200
Hagen Paul Pfeifer <hagen@jauu.net> wrote:
> * Stephen Hemminger | 2010-06-19 22:16:11 [-0700]:
>
> >I don't think this works because it breaks ABI compatibility for applications tha
> >use older versions.
>
> Are you sure Stephen? It is a one-to-one mapping of the ABI but maybe it was
> too late yesterday ... ;-)
I was worried that the new (remapped codes) would overlap the ones from user
space. Maybe best to have separate structures for the userspace API and the
kernel optimized filter instructions. You could then do what ever transformations
you want there.
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Ben Hutchings @ 2010-06-20 15:19 UTC (permalink / raw)
To: Herbert Xu
Cc: Michael S. Tsirkin, Xin, Xiaohui, Stephen Hemminger,
netdev@vger.kernel.org, kvm@vger.kernel.org,
linux-kernel@vger.kernel.org, mingo@elte.hu, davem@davemloft.net,
jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620115926.GA31849@gondor.apana.org.au>
On Sun, 2010-06-20 at 21:59 +1000, Herbert Xu wrote:
> On Sun, Jun 20, 2010 at 02:47:19PM +0300, Michael S. Tsirkin wrote:
> >
> > Let's do this then. So far the virtio spec avoided making layout
> > assumptions, leaving guests lay out data as they see fit.
> > Isn't it possible to keep supporting this with zero copy for hardware
> > that can issue DMA at arbitrary addresses?
>
> I think you're mistaken with respect to what is being proposed.
> Raising 512 bytes isn't a hard constraint, it is merely an
> optimisation for Intel NICs because their PS mode can produce
> a head fragment of up to 512 bytes.
>
> If the guest didn't allocate 512 bytes it wouldn't be the end of
> the world, it'd just mean that we'd either copy whatever is in
> the head fragment, or we waste 4096-X bytes of memory where X
> is the number of bytes in the head.
If I understand correctly what this 'PS mode' is (I haven't seen the
documentation for it), it is a feature that Microsoft requested from
hardware vendors for use in Hyper-V. As a result, the SFC9000 family
and presumably other controllers also implement something similar.
Ben.
--
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: Distributed Switch Architecture(DSA)
From: Joakim Tjernlund @ 2010-06-20 14:41 UTC (permalink / raw)
To: Lennert Buytenhek; +Cc: netdev
In-Reply-To: <20100619185739.GQ14513@mail.wantstofly.org>
Lennert Buytenhek <buytenh@wantstofly.org> wrote on 2010/06/19 20:57:39:
>
> On Sat, Jun 19, 2010 at 08:48:31PM +0200, Joakim Tjernlund wrote:
>
> > > > > > Not sure how one would express whether VLAN tags should be stripped
> > > > > > off or not when egressing the HW switch's physical port.
> > > > >
> > > > > If you transmit a packet onto 'lan', it will be sent to the switch chip
> > > > > with an "untagged" DSA tag. If you transmit a packet onto 'lan.123',
> > > > > it will be sent to the switch chip with a "tagged" DSA tag. See
> > > > > net/dsa/tag_dsa.c for details.
> > > >
> > > > Ah, now I get it, thanks.
> > > > However, how does this work for LAN to LAN pkgs? LAN1 and LAN2 could be
> > > > in the same VLAN but one is implicit(port) VLAN and the
> > > > other is explicit.
> > >
> > > If you tell the HW switch to forward these packets, they will never
> > > appear at the CPU interface, so the DSA tagging/untagging doesn't enter
> > > the picture.
> >
> > "tell the HW switch"? Doesn't DSA do that already?
>
> Not in its current iteration, as I've explained in previous emails.
Sorry, I didn't quite get that.
>
>
> > If not, what is the point of DSA then if it doesn't use the native
> > forwarding capabilities of the HW switch?
>
> The point is and always was to provide a framework for proper integration
> of hardware switch chips into the Linux kernel. This framework doesn't
> become useless just because it doesn't already support every single
> hardware feature at this point.
Right, sorry if I sounded a bit harsh.
So DSA currently does a very minimal config of the HW switch to get
things going.
If you want to do something more fancy one has to
add a control plane to DSA which would possibly talk
to a user space app. Is that correct?
>
>
> > > > How do I config the HW switch to do that?
> > >
> > > Tell the switch that the vlan is native on one of the ports but not on
> > > the other. It's been a while since I looked at the chip docs but there
> > > are ways of doing this.
> >
> > The current DSA impl. does not support this? There should be some
> > way to manage this within the DSA framework.
>
> Have you even tried the DSA code?
Not yet and I don't have any MV HW either :(
^ permalink raw reply
* Re: [PATCH] Driver-core: Always create class directories fixing the broken network drivers.
From: Kay Sievers @ 2010-06-20 13:37 UTC (permalink / raw)
To: Eric W. Biederman; +Cc: Johannes Berg, Greg KH, netdev
In-Reply-To: <m1hbkx6g6p.fsf@fess.ebiederm.org>
On Sun, Jun 20, 2010 at 14:29, Eric W. Biederman <ebiederm@xmission.com> wrote:
> Kay Sievers <kay.sievers@vrfy.org> writes:
>> On Sun, Jun 20, 2010 at 13:33, Johannes Berg <johannes@sipsolutions.net> wrote:
>>> On Sun, 2010-06-20 at 12:52 +0200, Kay Sievers wrote:
>>>
>>>> As mentioned earlier, It's pretty fragile to change things in this
>>>> area, and I prefer the broken network driver-core interactions to be
>>>> fixed instead - even when they are more complicated.
>>>
>>> Can you _please_ offer a proper way to fix it then?
>>
>> Sorry, I have no real experience with the issues created by the
>> assumption that network driver need to be able to get unloaded while
>> in use. That's very special, always requires a
>> compiled-into-the-kernel part of the subsystem, and makes it hard to
>> work with, as we can not use any of the usual core infrastructure to
>> solve that.
>
> So please look at https://bugzilla.kernel.org/show_bug.cgi?id=16215
>
> That simply creates and destroys the network device as things come
> and go.
I'm still not sure, any help here would be appreciated.
> I think the bnep case is much more serious because it is real hardware
> not a testing simulation, and it is the second instance of this.
>
> Calling the change broken when I can boot up and run X in that
> configuration just fine is a vast overstatement.
Oh, I seriously would love this rule - it would make my work so much
easier. But I need to make it totally clear: "Adding intermediate
directories into 'input' sysfs it absolutely broken, regardless if
your box comes up or not. :)
X is using udev, and udev aggressively hides these details and forbids
matching such details, but many other tools which read sysfs directly,
including ones using the conceptually broken 'device' symlink will for
sure break with such changes.
> Especially
> when you don't acknowledge that the device layer is broken.
Stacking devices from different classes is broken, and not a direct
problem of the core. It is just not supported. The core might just
need to refuse that in the first place, but that's a different issue.
> I will agree that insane amounts of backwards compatibility are a good
> idea. So I will cook up a version of my patch that adds a hack to the
> device layer to only apply this change to devices of class net.
>
> That should save let us postpone the architectural dreams for another
> day.
It's not a dream, it needs to be fixed where it is used. We can not
allow to stack classes.
Kay
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Michael S. Tsirkin @ 2010-06-20 12:48 UTC (permalink / raw)
To: Herbert Xu
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620115926.GA31849@gondor.apana.org.au>
On Sun, Jun 20, 2010 at 09:59:26PM +1000, Herbert Xu wrote:
> On Sun, Jun 20, 2010 at 02:47:19PM +0300, Michael S. Tsirkin wrote:
> >
> > Let's do this then. So far the virtio spec avoided making layout
> > assumptions, leaving guests lay out data as they see fit.
> > Isn't it possible to keep supporting this with zero copy for hardware
> > that can issue DMA at arbitrary addresses?
>
> I think you're mistaken with respect to what is being proposed.
> Raising 512 bytes isn't a hard constraint, it is merely an
> optimisation for Intel NICs because their PS mode can produce
> a head fragment of up to 512 bytes.
>
Thanks for the clarification. So what is discussed here is
the API changes that will enable this optimization?
Of couse, it makes sense to consider this to try and avoid code churn
in the future.
As a side note, I hope to see a basic zero copy implementation with
GSO/GRO that beats copy in host convincingly before work is started on
further optimizations, though.
> If the guest didn't allocate 512 bytes it wouldn't be the end of
> the world, it'd just mean that we'd either copy whatever is in
> the head fragment,
I don't know how much will copying the head cost.
> or we waste 4096-X bytes of memory where X
> is the number of bytes in the head.
This seems mostly harmless - and guest can always do a copy internally
to save memory, correct?
Note also that we lock a full page to allow DMA, anyway.
> Cheers,
> --
> Visit Openswan at http://www.openswan.org/
> Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
> Home Page: http://gondor.apana.org.au/~herbert/
> PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* [PATCH] Driver-core: Always create network class directories in get_device_parent.
From: Eric W. Biederman @ 2010-06-20 12:46 UTC (permalink / raw)
To: Greg KH; +Cc: Johannes Berg, netdev, Kay Sievers
In-Reply-To: <AANLkTikZAKZA9mfb7RjaqNFnX8gCM3A9PTfH350botL8@mail.gmail.com>
In get_device_parent there was added check to not add a class
directory when a class device was put under another class device. The
check was put in place as a just in case measure to not break old
userspace if any existing code happened to depend on it. Devices in
the input subsystem are affected by this code path so there is a
reasonable chance that some piece of user space will break if we just
remove this kludge.
At the same time there are at least two network drivers that have
potential unnecessary namespace conflicts because class directories
have not been created for their network devices.
With the introduction of tagged sysfs directories for properly
handling network namespace support this omission in creating the class
directories went from a bad thing in terms of namespace pollution, to
actually breaking device_remove.
Currently there are two reported network device drivers that break
because the class directory was not created by the device layer. The
usb bnep driver and the mac80211_hwsim driver.
Every solution proposed changes the sysfs layout for the affected
devices, and thus has the potential to break userspace.
Since we are changing the sysfs layout anyway, and since we are now
talking about several devices all with the same problem, all caused by
the same over convservative bit of code. Let's fix the device layer
for network devices.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
---
drivers/base/core.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 9630fbd..ffb8443 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -673,7 +673,7 @@ static struct kobject *get_device_parent(struct device *dev,
*/
if (parent == NULL)
parent_kobj = virtual_device_parent(dev);
- else if (parent->class)
+ else if (parent->class && (strcmp(dev->class->name, "net") != 0))
return &parent->kobj;
else
parent_kobj = &parent->kobj;
--
1.6.5.2.143.g8cc62
^ permalink raw reply related
* Re: [PATCH] Driver-core: Always create class directories fixing the broken network drivers.
From: Eric W. Biederman @ 2010-06-20 12:29 UTC (permalink / raw)
To: Kay Sievers; +Cc: Johannes Berg, Greg KH, netdev
In-Reply-To: <AANLkTikZAKZA9mfb7RjaqNFnX8gCM3A9PTfH350botL8@mail.gmail.com>
Kay Sievers <kay.sievers@vrfy.org> writes:
> On Sun, Jun 20, 2010 at 13:33, Johannes Berg <johannes@sipsolutions.net> wrote:
>> On Sun, 2010-06-20 at 12:52 +0200, Kay Sievers wrote:
>>
>>> As mentioned earlier, It's pretty fragile to change things in this
>>> area, and I prefer the broken network driver-core interactions to be
>>> fixed instead - even when they are more complicated.
>>
>> Can you _please_ offer a proper way to fix it then?
>
> Sorry, I have no real experience with the issues created by the
> assumption that network driver need to be able to get unloaded while
> in use. That's very special, always requires a
> compiled-into-the-kernel part of the subsystem, and makes it hard to
> work with, as we can not use any of the usual core infrastructure to
> solve that.
So please look at https://bugzilla.kernel.org/show_bug.cgi?id=16215
That simply creates and destroys the network device as things come
and go.
I think the bnep case is much more serious because it is real hardware
not a testing simulation, and it is the second instance of this.
Calling the change broken when I can boot up and run X in that
configuration just fine is a vast overstatement. Especially
when you don't acknowledge that the device layer is broken.
I will agree that insane amounts of backwards compatibility are a good
idea. So I will cook up a version of my patch that adds a hack to the
device layer to only apply this change to devices of class net.
That should save let us postpone the architectural dreams for another
day.
Eric
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Herbert Xu @ 2010-06-20 11:59 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620114719.GC5285@redhat.com>
On Sun, Jun 20, 2010 at 02:47:19PM +0300, Michael S. Tsirkin wrote:
>
> Let's do this then. So far the virtio spec avoided making layout
> assumptions, leaving guests lay out data as they see fit.
> Isn't it possible to keep supporting this with zero copy for hardware
> that can issue DMA at arbitrary addresses?
I think you're mistaken with respect to what is being proposed.
Raising 512 bytes isn't a hard constraint, it is merely an
optimisation for Intel NICs because their PS mode can produce
a head fragment of up to 512 bytes.
If the guest didn't allocate 512 bytes it wouldn't be the end of
the world, it'd just mean that we'd either copy whatever is in
the head fragment, or we waste 4096-X bytes of memory where X
is the number of bytes in the head.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Michael S. Tsirkin @ 2010-06-20 11:47 UTC (permalink / raw)
To: Herbert Xu
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620113609.GA31693@gondor.apana.org.au>
On Sun, Jun 20, 2010 at 09:36:09PM +1000, Herbert Xu wrote:
> On Sun, Jun 20, 2010 at 02:11:24PM +0300, Michael S. Tsirkin wrote:
> >
> > Rather than modifying all guests, it seems much easier not to assume
> > specific buffer layout in host. Copying network header around seems a
> > small cost.
>
> Well sure we can debate the specifics of this implementation detail.
Let's do this then. So far the virtio spec avoided making layout
assumptions, leaving guests lay out data as they see fit.
Isn't it possible to keep supporting this with zero copy for hardware
that can issue DMA at arbitrary addresses?
> However, the fact that virtio_net doesn't support feature renegotiation
> on live migration is not a valid reason against this.
>
> Cheers,
--
MST
^ permalink raw reply
* Re: [PATCH] Driver-core: Always create class directories fixing the broken network drivers.
From: Kay Sievers @ 2010-06-20 11:46 UTC (permalink / raw)
To: Johannes Berg; +Cc: Eric W. Biederman, Greg KH, netdev
In-Reply-To: <1277033628.3642.1.camel@jlt3.sipsolutions.net>
On Sun, Jun 20, 2010 at 13:33, Johannes Berg <johannes@sipsolutions.net> wrote:
> On Sun, 2010-06-20 at 12:52 +0200, Kay Sievers wrote:
>
>> As mentioned earlier, It's pretty fragile to change things in this
>> area, and I prefer the broken network driver-core interactions to be
>> fixed instead - even when they are more complicated.
>
> Can you _please_ offer a proper way to fix it then?
Sorry, I have no real experience with the issues created by the
assumption that network driver need to be able to get unloaded while
in use. That's very special, always requires a
compiled-into-the-kernel part of the subsystem, and makes it hard to
work with, as we can not use any of the usual core infrastructure to
solve that.
The only real simple thing that works is splitting the module in two
modules, which isn't really something I would propose.
Maybe the wait-for in the module-exit like your recent mail suggests
works, but I did not try that. Otherwise we can solve this by changing
the net driver and by adding some needed stuff to the core to allow
in-core bus device cleanup.
The class device hierarchy should be removed for proper network
namespace support, it's nothing we properly support with the current
core code. We better don't fiddle around with stuff nobody really
knows what it breaks. Just like I ran into the 'input' stuff now,
which was a really simple case to find.
Kay
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Herbert Xu @ 2010-06-20 11:36 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620111124.GB5285@redhat.com>
On Sun, Jun 20, 2010 at 02:11:24PM +0300, Michael S. Tsirkin wrote:
>
> Rather than modifying all guests, it seems much easier not to assume
> specific buffer layout in host. Copying network header around seems a
> small cost.
Well sure we can debate the specifics of this implementation detail.
However, the fact that virtio_net doesn't support feature renegotiation
on live migration is not a valid reason against this.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [PATCH] Driver-core: Always create class directories fixing the broken network drivers.
From: Johannes Berg @ 2010-06-20 11:33 UTC (permalink / raw)
To: Kay Sievers; +Cc: Eric W. Biederman, Greg KH, netdev
In-Reply-To: <AANLkTilVHF0dITpB3yJYaCnGPonhdl7NWnb-P12RAFTs@mail.gmail.com>
On Sun, 2010-06-20 at 12:52 +0200, Kay Sievers wrote:
> As mentioned earlier, It's pretty fragile to change things in this
> area, and I prefer the broken network driver-core interactions to be
> fixed instead - even when they are more complicated.
Can you _please_ offer a proper way to fix it then?
johannes
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Michael S. Tsirkin @ 2010-06-20 11:11 UTC (permalink / raw)
To: Herbert Xu
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620110254.GA31484@gondor.apana.org.au>
On Sun, Jun 20, 2010 at 09:02:54PM +1000, Herbert Xu wrote:
> On Sun, Jun 20, 2010 at 01:39:09PM +0300, Michael S. Tsirkin wrote:
> >
> > > It's just like migrating from a backend that supports TSO
> > > to one that doesn't.
> >
> > Exactly. We don't support such migration.
>
> Well that's something that has to be addressed in the virtio_net.
Rather than modifying all guests, it seems much easier not to assume
specific buffer layout in host. Copying network header around seems a
small cost.
> Cheers,
> --
> Visit Openswan at http://www.openswan.org/
> Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
> Home Page: http://gondor.apana.org.au/~herbert/
> PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Herbert Xu @ 2010-06-20 11:02 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620103909.GA5285@redhat.com>
On Sun, Jun 20, 2010 at 01:39:09PM +0300, Michael S. Tsirkin wrote:
>
> > It's just like migrating from a backend that supports TSO
> > to one that doesn't.
>
> Exactly. We don't support such migration.
Well that's something that has to be addressed in the virtio_net.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [PATCH] Driver-core: Always create class directories fixing the broken network drivers.
From: Kay Sievers @ 2010-06-20 10:52 UTC (permalink / raw)
To: Eric W. Biederman; +Cc: Greg KH, Johannes Berg, netdev
In-Reply-To: <m139wiuswa.fsf_-_@fess.ebiederm.org>
On Sun, Jun 20, 2010 at 08:20, Eric W. Biederman <ebiederm@xmission.com> wrote:
>
> In get_device_parent there is a check to not add a class directory
> when a class device was put under another class device. The check was
> put in place as a just in case measure to not break old userspace if
> any existing code happened to depend on it. Currently the only known
> way that we get a class device under a class device is due to the
> rearrangement of devices that happened when the new sysfs layout was
> introduced.
>
> With the introduction of tagged sysfs directories for properly
> handling network namespace support this omission in creating the class
> directories went from a bad thing in terms of namespace pollution, to
> actually breaking device_remove.
>
> Currently there are two reported network device drivers that break
> because the class directory was not created by the device layer. The
> usb bnep driver and the mac80211_hwsim driver.
>
> Every solution proposed changes the sysfs layout for the affected
> devices, and thus has the potential to break userspace.
>
> Since we are changing the sysfs layout anyway, and since we are now
> talking about several devices all with the same problem, all caused by
> the same over conservative bit of code. Let's kill that bit of code.
>
> There have been other proposals to fix this but they all have been
> more complicated, and none of them have actually resulted in working
> code.
>
> Any userspace that works with both the old and the new sysfs layouts
> should not be affected by this change, and even if someone depends
> on it we are talking a very small number of drivers overall that
> are affected.
>
> My apologoies for not fully catching this hole in the logic the
> when this code was originally added.
We can not do this. Simply comparing the sysfs tree before and after
shows that it breaks 'input'. inputX and mouseX are now spearated by a
subdirectory, which is wrong.
As mentioned earlier, It's pretty fragile to change things in this
area, and I prefer the broken network driver-core interactions to be
fixed instead - even when they are more complicated.
Thanks,
Kay
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Michael S. Tsirkin @ 2010-06-20 10:39 UTC (permalink / raw)
To: Herbert Xu
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620103235.GA31284@gondor.apana.org.au>
On Sun, Jun 20, 2010 at 08:32:35PM +1000, Herbert Xu wrote:
> On Sun, Jun 20, 2010 at 01:06:32PM +0300, Michael S. Tsirkin wrote:
> >
> > Changing the guest virtio to match the backend is a problem,
> > this breaks migration etc.
>
> As long as it's done in a backwards compatible way it should be
> fine.
Possibly, but to me the need to do this implies that
we'll need another change with different hardware at the backend.
> It's just like migrating from a backend that supports TSO
> to one that doesn't.
>
> Cheers,
Exactly. We don't support such migration.
> --
> Visit Openswan at http://www.openswan.org/
> Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
> Home Page: http://gondor.apana.org.au/~herbert/
> PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Herbert Xu @ 2010-06-20 10:32 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Xin, Xiaohui, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <20100620100631.GB4578@redhat.com>
On Sun, Jun 20, 2010 at 01:06:32PM +0300, Michael S. Tsirkin wrote:
>
> Changing the guest virtio to match the backend is a problem,
> this breaks migration etc.
As long as it's done in a backwards compatible way it should be
fine. It's just like migrating from a backend that supports TSO
to one that doesn't.
Cheers,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
From: Michael S. Tsirkin @ 2010-06-20 10:06 UTC (permalink / raw)
To: Xin, Xiaohui
Cc: Herbert Xu, Stephen Hemminger, netdev@vger.kernel.org,
kvm@vger.kernel.org, linux-kernel@vger.kernel.org, mingo@elte.hu,
davem@davemloft.net, jdike@linux.intel.com, Rusty Russell
In-Reply-To: <F2E9EB7348B8264F86B6AB8151CE2D7915089FE573@shsmsx502.ccr.corp.intel.com>
On Fri, Jun 18, 2010 at 03:14:18PM +0800, Xin, Xiaohui wrote:
> >-----Original Message-----
> >From: Herbert Xu [mailto:herbert@gondor.apana.org.au]
> >Sent: Friday, June 18, 2010 1:59 PM
> >To: Xin, Xiaohui
> >Cc: Stephen Hemminger; netdev@vger.kernel.org; kvm@vger.kernel.org;
> >linux-kernel@vger.kernel.org; mst@redhat.com; mingo@elte.hu; davem@davemloft.net;
> >jdike@linux.intel.com; Rusty Russell
> >Subject: Re: [RFC PATCH v7 01/19] Add a new structure for skb buffer from external.
> >
> >On Fri, Jun 18, 2010 at 01:26:49PM +0800, Xin, Xiaohui wrote:
> >>
> >> Herbert,
> >> I have questions about the idea above:
> >> 1) Since netdev_alloc_skb() is still there, and we only modify alloc_page(),
> >> then we don't need napi_gro_frags() any more, the driver's original receiving
> >> function is ok. Right?
> >
> >Well I was actually thinking about converting all drivers that
> >need this to napi_gro_frags. But now that you mention it, yes
> >we could still keep the old interface to minimise the work.
> >
> >> 2) Is napi_gro_frags() only suitable for TCP protocol packet?
> >> I have done a small test for ixgbe driver to let it only allocate paged buffers
> >> and found kernel hangs when napi_gro_frags() receives an ARP packet.
> >
> >It should work with any packet. In fact, I'm pretty sure the
> >other drivers (e.g., cxgb3) use that interface for all packets.
> >
> Thanks for the verification. By the way, does that mean that nearly all drivers can use the
> same napi_gro_frags() to receive buffers though currently each driver has it's own receiving
> function?
>
> >> 3) As I have mentioned above, with this idea, netdev_alloc_skb() will allocate
> >> as usual, the data pointed by skb->data will be copied into the first guest buffer.
> >> That means we should reserve sufficient room in guest buffer. For PS mode
> >> supported driver (for example ixgbe), the room will be more than 128. After 128bytes,
> >> we will put the first frag data. Look into virtio-net.c the function page_to_skb()
> >> and receive_mergeable(), that means we should modify guest virtio-net driver to
> >> compute the offset as the parameter for skb_set_frag().
> >>
> >> How do you think about this? Attached is a patch to how to modify the guest driver.
> >> I reserve 512 bytes as an example, and transfer the header len of the skb in hdr->hdr_len.
> >
> >Expanding the buffer size to 512 bytes to accomodate PS mode
> >looks reasonable to me.
> >
> >However, I don't think we should increase the copy threshold to
> >512 bytes at the same time. I don't have any figures myself but
> >I think if we are to make such a change it should be a separate
> >one and come with supporting numbers.
> >
> Let me have a look to see if I can retain the copy threshold as 128 bytes
> and copy the header data safely.
Changing the guest virtio to match the backend is a problem,
this breaks migration etc.
> >Cheers,
> >--
> >Visit Openswan at http://www.openswan.org/
> >Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
> >Home Page: http://gondor.apana.org.au/~herbert/
> >PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: [PATCH] net: optimize Berkeley Packet Filter (BPF) processing
From: Hagen Paul Pfeifer @ 2010-06-20 9:50 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: netdev, davem
In-Reply-To: <20100619221611.784f7dbc@nehalam>
* Stephen Hemminger | 2010-06-19 22:16:11 [-0700]:
>I don't think this works because it breaks ABI compatibility for applications tha
>use older versions.
Are you sure Stephen? It is a one-to-one mapping of the ABI but maybe it was
too late yesterday ... ;-)
--
Hagen Paul Pfeifer <hagen@jauu.net> || http://jauu.net/
Telephone: +49 174 5455209 || Key Id: 0x98350C22
Key Fingerprint: 490F 557B 6C48 6D7E 5706 2EA2 4A22 8D45 9835 0C22
^ permalink raw reply
* Re: [net-next-2.6 PATCH 1/8] e1000e: cleanup ethtool loopback setup code
From: Jeff Kirsher @ 2010-06-20 7:32 UTC (permalink / raw)
To: David Miller; +Cc: netdev, gospo, bphilips, bruce.w.allan
In-Reply-To: <20100618.221512.102550313.davem@davemloft.net>
On Fri, Jun 18, 2010 at 22:15, David Miller <davem@davemloft.net> wrote:
>
> I've applied this series however:
>
> 1) Please address Ben's concerns about turning EEE on by default
> given that standardization is not complete yet.
>
> 2) I hate module parameters, I'd rather you create a new ethtool
> feature bit and thus allow the setting to be modified at run
> time. Please create a new ethtool control flag, and remove
> this module option.
>
> Thanks.
> --
Thank you Dave. I know that Bruce (and company) are actively looking
into your first point. I would think that he would have either a
response or update on the status come Monday/Tuesday.
Regarding option 2, understood and I will work out with Bruce on when
and who will do the work. I will be taking a look at it tomorrow
(later today), if I am not able to come up with a patch in the near
future. I will return to working on it by Wednesday.
--
Cheers,
Jeff
^ permalink raw reply
* [PATCH] Driver-core: Always create class directories fixing the broken network drivers.
From: Eric W. Biederman @ 2010-06-20 6:20 UTC (permalink / raw)
To: Greg KH; +Cc: Johannes Berg, netdev, Kay Sievers
In-Reply-To: <AANLkTin2sci1gSmwx-tjazYGFAcUNRKhJFI7Bt0KvVe3@mail.gmail.com>
In get_device_parent there is a check to not add a class directory
when a class device was put under another class device. The check was
put in place as a just in case measure to not break old userspace if
any existing code happened to depend on it. Currently the only known
way that we get a class device under a class device is due to the
rearrangement of devices that happened when the new sysfs layout was
introduced.
With the introduction of tagged sysfs directories for properly
handling network namespace support this omission in creating the class
directories went from a bad thing in terms of namespace pollution, to
actually breaking device_remove.
Currently there are two reported network device drivers that break
because the class directory was not created by the device layer. The
usb bnep driver and the mac80211_hwsim driver.
Every solution proposed changes the sysfs layout for the affected
devices, and thus has the potential to break userspace.
Since we are changing the sysfs layout anyway, and since we are now
talking about several devices all with the same problem, all caused by
the same over conservative bit of code. Let's kill that bit of code.
There have been other proposals to fix this but they all have been
more complicated, and none of them have actually resulted in working
code.
Any userspace that works with both the old and the new sysfs layouts
should not be affected by this change, and even if someone depends
on it we are talking a very small number of drivers overall that
are affected.
My apologoies for not fully catching this hole in the logic the
when this code was originally added.
Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
---
drivers/base/core.c | 2 --
1 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/drivers/base/core.c b/drivers/base/core.c
index 9630fbd..7b1c4d4 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -673,8 +673,6 @@ static struct kobject *get_device_parent(struct device *dev,
*/
if (parent == NULL)
parent_kobj = virtual_device_parent(dev);
- else if (parent->class)
- return &parent->kobj;
else
parent_kobj = &parent->kobj;
--
1.6.5.2.143.g8cc62
^ permalink raw reply related
* Re: [PATCH] net: optimize Berkeley Packet Filter (BPF) processing
From: Stephen Hemminger @ 2010-06-20 5:16 UTC (permalink / raw)
To: Hagen Paul Pfeifer; +Cc: netdev, davem
In-Reply-To: <1277003136-5522-1-git-send-email-hagen@jauu.net>
On Sun, 20 Jun 2010 05:05:36 +0200
Hagen Paul Pfeifer <hagen@jauu.net> wrote:
> Gcc is currenlty not in the ability to optimize the switch statement in
> sk_run_filter() because of dense case labels. This patch replace the
> OR'd labels with ordered sequenced case labels. The sk_chk_filter()
> function is modified to patch/replace the original OPCODES in a
> ordered but equivalent form. gcc is now in the ability to transform the
> switch statement in sk_run_filter into a jump table of complexity O(1).
>
> Until this patch gcc generates a sequence of conditional branches (O(n) of 567
> byte .text segment size (arch x86_64):
I don't think this works because it breaks ABI compatibility for applications tha
use older versions.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox