On Iau, 2005-04-21 at 06:54, Albert Lee wrote:
> A particular revision of the HPT372N oopses hpt366 consistently. It's a
> regression caused by Alan's changes in 2.6.9 to support the HPT372N using
> only PLL timings. The driver works correctly in prior versions, where the the
> PCI clock is used
Hi,
These are the sizes of rc2 and rc3 patches
# ls -la patch-2.6.12*
-rw-r--r-- 1 root src 18011382 Apr 4 18:50 patch-2.6.12-rc2
-rw-r--r-- 1 root src 19979854 Apr 21 02:29 patch-2.6.12-rc3
Let us make an incremental patch from rc2 to rc3
# interdiff patch-2.6.12-rc2 patch-2.6.12-rc3 >x
Le
On Thu, 2005-04-21 at 14:22 +1000, David Woodhouse wrote:
> As of some time in the fairly near future, the [EMAIL PROTECTED] mailing
> list will be carrying real commits from Linus' live git repository, instead
> of just testing patches. Have fun.
>
with BK this was not possible, but could we pl
On 4/20/05, Adrian Bunk <[EMAIL PROTECTED]> wrote:
> The times when tricky goto's produced better codes are long gone.
>
> This patch should express the same in a better way, please check whether
> I made any mistake.
>
By the way, it solves compile errors with gcc-4:
a lot of
drivers/net/hamrad
On Thu, Apr 21, 2005 at 11:58:12AM +0200, Geert Uytterhoeven wrote:
> sun3_NCR5380.c still uses the following:
>
> - SCSI_ABORT_SUCCESS
> - SCSI_ABORT_ERROR
> - SCSI_ABORT_SNOOZE
> - SCSI_ABORT_BUSY
> - SCSI_ABORT_NOT_RUNNING
> - SCSI_RESET_SUCCESS
> - SCSI_RESET_BUS_RESET
>
> causi
On Tue, Apr 19, 2005 at 10:20:38PM +0200, Domen Puncer wrote:
> > Please keep using __inline__, not inline.
>
> Why?
>
> Couldn't find any threads about this, and even SubmittingPatches has:
> "'static inline' is preferred over 'static __inline__'..."
Unlike inline __inline__ will be recogniced
On Tue, Apr 19, 2005 at 03:24:00PM -0400, Rik van Riel wrote:
> The current development model seems to go much smoother than
> anything I've seen before.
It violates conventional wisdom and that psychological thing is the lion
share of why some people feel uneasy about it.
Dealing with the paral
On Tue, Apr 19, 2005 at 09:24:17AM -0400, Lennart Sorensen wrote:
> If you change it now, how many tools would break?
>
> Maybe if you can list what statistics you think should be common to all
> systems, that could be presented in another file that is always the same
> format on each architectur
On Wed, 20 Apr 2005, Linux Kernel Mailing List wrote:
> tree 932c6f9689fd08a7a9d689cfbec8682ccde8175d
> parent 84011ae88da62a20b3ae7b48e2ae3b1ef0fc810a
> author <[EMAIL PROTECTED]> Mon, 11 Apr 2005 08:19:25 -0500
> committer James Bottomley <[EMAIL PROTECTED]> Sun, 17 Apr 2005 06:14:52 -0500
>
> [
Hi!
A small update.
Patching mm/filemap.c is not necessary in order to get the improved
performance!
It's sufficient to remove roundup_pow_of_two from |get_init_ra_size ...
So a simple one-liner changes to picture dramatically.
But why ?!?!?
roundup_pow_of_two() uses fls() and ia64 has buggy fls()
Hello,
'make prepare ARCH=um' spits out some serious errors.
'make ARCH=um' works though - just takes ways longer ...
Here are the 3 commands to reproduce plus the errornous output:
-
Stone:/usr/src/linux-2.6.11.7 # uname -a
Linux Stone 2.
On Wed, 20 Apr 2005 14:35:10 +0200 Arjan van de Ven wrote:
>
> > diff -puN drivers/base/node.c~numa_hp_base drivers/base/node.c
> > --- linux-2.6.12-rc2-mm3/drivers/base/node.c~numa_hp_base 2005-04-14
> > 20:49:37.0 +0900
> > +++ linux-2.6.12-rc2-mm3-kei/drivers/base/node.c2005-04-1
On Thu, 21 Apr 2005, Jan Dittmer wrote:
> Linus Torvalds wrote:
> > Geert Uytterhoeven:
> > [PATCH] M68k: Update defconfigs for 2.6.11
> > [PATCH] M68k: Update defconfigs for 2.6.12-rc2
>
> Why do I still get this error when trying to cross-compile for m68k?
Because to build m68k kernels,
Mikael Andersson wrote:
> During heavy io-load a lockup occurs that appears to prevent any disk
> output from taking place. fs is reiserfs on two device-mapper mirrored
> 200G maxtor disks. After the lockup occurs you can to things like 'ls',
> but echo > test.txt will hang.
fs is now ext3
>
> A
Linus Torvalds wrote:
> Geert Uytterhoeven:
> [PATCH] M68k: Update defconfigs for 2.6.11
> [PATCH] M68k: Update defconfigs for 2.6.12-rc2
Why do I still get this error when trying to cross-compile for m68k?
toolchain:
Reading specs from
/usr/local/m68k-uclinux-tools/lib/gcc/m68k-uclinux
W1: support for automatic family drivers loading via hotplug:
- allow family drivers support list of families;
- export supported families through MODULE_DEVICE_TABLE.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
drivers/w1/w1.c |6 --
drivers/w1/w1.h
On Wed, Apr 20, 2005 at 11:40:16PM -0400, Steven Rostedt wrote:
> Is 11 jiffies correct for 10ms?
Consider the 1 jiffy case. How long does waiting one jiffy actually wait?
j=01 2
+--+--+--> t
A B C D
If you start timing one jiffy from A
> [EMAIL PROTECTED]:
> [PATCH] zfcp: convert to compat_ioctl
This does not seem to compile anymore with defconfig:
CC drivers/s390/scsi/zfcp_aux.o
/usr/src/ctest/rc/kernel/drivers/s390/scsi/zfcp_aux.c:63: warning:
initialization from incompatible pointer type
/usr/src/ctest/rc/kernel/
On Thu, 2005-04-21 at 17:20 +1000, Nick Piggin wrote:
> On Thu, 2005-04-21 at 08:10 +0100, Anton Altaparmakov wrote:
> > And one more thing...
> >
> > On Thu, 2005-04-21 at 08:01 +0100, Anton Altaparmakov wrote:
> > > On Thu, 21 Apr 2005, Nick Piggin wrote:
> > > > ... I somehow didn't send it to
2005/4/21, Ingo Molnar <[EMAIL PROTECTED]>:
>
> i have released the -V0.7.46-00 Real-Time Preemption patch, which can be
> downloaded from the usual place:
>
> http://redhat.com/~mingo/realtime-preempt/
>
> this is a merge to 2.6.12-rc3, plus the 'ping localhost' fix from
> [EMAIL PROTECTED]
>
W1: allow changing w1 module parameters through sysfs, add parameter
descriptions and document them in Documentation/kernel-parameters.txt
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
Documentation/kernel-parameters.txt | 19 ---
drivers/w1/w1.c
A small update.
Patching mm/filemap.c is not necessary in order to get the improved
performance!
It's sufficient to remove roundup_pow_of_two from |get_init_ra_size ...
So a simple one-liner changes to picture dramatically.
But why ?!?!?
Andreas
|
jmerkey wrote:
For 3Ware, you need to chage the
Ahem.. Kmail just refuses send this on inline... Sorry.
--
Dmitry
W1: implement W1 bus hotplug handler. Slave devices will define
FID (family ID) end SN (serial number) environment variables.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c | 51
On Wed, 2005-04-20 at 17:59 -0700, Linus Torvalds wrote:
> And for the crazy people, the git archive on kernel.org is up and running
> under /pub/scm/linux/kernel/git/torvalds/linux-2.6.git. For the
> adventurous of you, the name of the 2.6.12-rc3 release is a very nice and
> readable:
>
>
Tony Lindgren wrote:
> * Pavel Machek <[EMAIL PROTECTED]> [050419 14:10]:
>>Hi!
>>
>>>The machine is a Pentium M 2.00 GHz, supporting C0-C4 processor power states.
>>>The machine run at 2.00 GHz all the time.
>>..
>>>_passing bm_history=0x (default) to processor module:_
>>>
>>>Average curr
W1: convert family into proper device-model drivers:
- embed driver structure into w1_family and register with the
driver core;
- do not try to manually bind slaves to familes, leave it to
the driver core;
- fold w1_family.c into w1.c
Signed-off-by: Dmitry Torokhov <[EMAIL
i have released the -V0.7.46-00 Real-Time Preemption patch, which can be
downloaded from the usual place:
http://redhat.com/~mingo/realtime-preempt/
this is a merge to 2.6.12-rc3, plus the 'ping localhost' fix from
[EMAIL PROTECTED]
there are still some unsolved slowdowns probably related
W1: clean-up family implementation:
- get rid of w1_family_ops and template attributes in w1_slave
structure and have family drivers create necessary attributes
themselves. There are too many different devices using 1-Wire
interface and it is impossible to fit them all into si
W1: more master attributes changes:
- rename timeout parameter/attribute to scan_interval to better
reflect its purpose;
- make scan_timeout be a per-device attribute and allow changing
it from userspace via sysfs;
- allow changing max_slave_count it from userspace as well.
W1: get rid of unneeded master device attributes:
- 'pointer' and 'attempts' are meaningless for userspace;
- information provided by 'slaves' and 'slave_count' can be
gathered from other sysfs bits;
- w1_slave_found has to be rearranged now that slave_count
field is gone.
W1: clean-up slave device implementation:
- get rid of separate refcount, rely on driver model to
enforce lifetime rules;
- pin w1 module until slave device is registered with sysfs
to make sure W1 core stays loaded.
- drop 'name' attribute as we already have it in bus_id.
W1: add slave_ttl attribute to w1 masters.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c | 24
1 files changed, 24 insertions(+)
Index: dtor/drivers/w1/w1.c
===
--- dtor.orig/drivers/w1/w1.c
W1: clean-up master device implementation:
- get rid of separate refcount, rely on driver model to
enforce lifetime rules;
- use atomic to generate unique master IDs;
- drop unused fields.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c | 263 ++
W1: clean-up master attribute implementation:
- drop unnecessary "w1_master" prefix from attribute names;
- do not acquire master->mutex when accessing attributes;
- move attribute code "closer" to the rest of master code.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c |
W1: move w1_search function to w1_io.c to be with the rest of IO code.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c| 87 --
w1.h|1
w1_io.c | 89 +++
W1: Drop control thread from w1 core, whatever it does can
also be done in the context of w1_remove_master_device.
Also, pin the module when registering new master device
to make sure that w1 core is not unloaded until last
device is gone. This simplifies logic a lot.
Signed-off-by
W1: drop custom-made hotplug over netlink notification
from w1 core. Standard hotplug mechanism should work
just fine (patch will follow).
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
drivers/w1/w1_netlink.c | 66 ---
drivers/w1/w1
W1: fold w1_int.c into w1.c - there is no point in artificially
separating code for master devices between 2 files.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
drivers/w1/w1_int.c| 181 -
drivers/w1/w1_int.h| 36 --
W1: cleanup bus operations code:
- have bus operatiions accept w1_master instead of unsigned long
and drop data field from w1_bus_master so the structure can be
statically allocated by driver implementing it;
- rename w1_bus_master to w1_bus_ops to avoid confusion with
w1_
W1: list handling cleanup. Most of the list_for_each_safe users
don't need *_safe variant, *_entry variant is better suited
in most places. Also, checking retrieved list element for
null is a bit pointless...
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c | 131
W1: drop owner field from w1_master and w1_slave structures.
Just having it there does not magically fixes lifetime
rules.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c |1 -
w1.h |2 --
w1_int.c |1 -
3 files changed, 4 deletions(-)
Index: dtor/drivers
W1: add 2 default attributes "family" and "serial" to slave
devices, every 1-Wire slave has them. Use attribute_group
to handle. The rest of slave attributes are left as is -
will be dealt with later.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c | 115
W1: use attribute_group to create master device attributes to
guarantee proper cleanup in case of failure. Also, hide
most of attribute define ugliness in macros.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
w1.c | 157 +-
w1: some formatting changes to bring the code in line with
CodingStyle guidelines.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
dscore.c |4 +---
w1.c | 21 +
2 files changed, 10 insertions(+), 15 deletions(-)
Index: dtor/drivers/w1/w1.c
=
W1: whitespace fixes.
Signed-off-by: Dmitry Torokhov <[EMAIL PROTECTED]>
---
Kconfig| 14 +++---
Makefile |2
ds_w1_bridge.c | 24 +-
dscore.c | 126 -
dscore.h |6 +-
matrox_w1.c| 10 ++
Hi,
I happened to take a look into drivers/w1 and found there bunch of thigs
that IMO should be changed:
- custom-made refcounting is racy
- lifetime rules need to be better enforced
- family framework is insufficient for many advanced w1 devices
- custom-made hotplug notification over netlink sh
* yangyi <[EMAIL PROTECTED]> wrote:
> Hi, Ingo
>
> For the option PREEMPT_RT, local communication latency is very very
> big, it is about 30 to 50 times as big as the option PREEMPT_NONE as
> far as local ping latency is concerned. Obviously, this should be
> fixed ASAP.
>
> This patch fixes
Hi guys,
One of Ben's patches ("ppc32: Fix cpufreq problems") went in 2.6.12-
rc3, but it depended on another patch that's still in -mm only:
add-suspend-method-to-cpufreq-core.patch
In addition to this, there's a third patch in -mm that fixes warnings
and line length to the previous patch, but
On Thu, 2005-04-21 at 08:10 +0100, Anton Altaparmakov wrote:
> And one more thing...
>
> On Thu, 2005-04-21 at 08:01 +0100, Anton Altaparmakov wrote:
> > On Thu, 21 Apr 2005, Nick Piggin wrote:
> > > ... I somehow didn't send it to Andrew last time.
> > >
> > > Fix a race where __block_prepare_wr
Ed L Cashin <[EMAIL PROTECTED]> wrote:
> +++ b/Documentation/aoe/aoe.txt 2005-04-20 11:42:20.0 -0400
> + When the aoe driver is a module, use
Is there any reason for this inconsistent behaviour?
> + /sys/module/aoe/parameters/aoe_iflist instead of
On Thu, 2005-04-21 at 08:01 +0100, Anton Altaparmakov wrote:
> Any reason why you left the goto out? It would be IMO much cleaner to
> remove the label "out" altogether and replace the single "goto out" with a
> "break" (which is fine since the goto happens inside the for loop
> immediately af
And one more thing...
On Thu, 2005-04-21 at 08:01 +0100, Anton Altaparmakov wrote:
> On Thu, 21 Apr 2005, Nick Piggin wrote:
> > ... I somehow didn't send it to Andrew last time.
> >
> > Fix a race where __block_prepare_write can leak out an in-flight
> > read against a bh if get_block returns an
Hi,
On Thu, 21 Apr 2005, Nick Piggin wrote:
> ... I somehow didn't send it to Andrew last time.
>
> Fix a race where __block_prepare_write can leak out an in-flight
> read against a bh if get_block returns an error. This can lead to
> the page becoming unlocked while the buffer is locked and the
201 - 253 of 253 matches
Mail list logo