> PS: No need to put a copy of the entire message
Apologies for the lapse in protocol.
> The point you're missing is that an "inactive" page is a free
> page that happens to have known clean data on it
I understand now where the inactive page count is coming from.
I don't understand why there
> From: Nick Piggin [mailto:[EMAIL PROTECTED]
> We had customers see similar incorrect OOM problems, so I sent in some
> patches merged after 2.6.16. Can you upgrade to latest kernel? (otherwise
> I guess backporting could be an option for you).
I will raise the question of moving the kernel forwa
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Reset infrastructure for re-IPL.
In case of re-IPL and diag308 doesn't work we have to reset all devices
manually and wait synchronously that each reset finished.
This patch adds the necessary infrastucture and the first exploiter of it.
Subsystems
From: Cornelia Huck <[EMAIL PROTECTED]>
[S390] cio: Make ccw_dev_id_is_equal() more robust.
Using memcmp to compare ccw_dev_id implies that the whole structure (incl.
padding) has always been completely initialized to sane values. Comparing
the structures field by field doesn't make such assumpti
From: Horst Hummel <[EMAIL PROTECTED]>
[S390] handle incorrect values when writing to dasd sysfs attributes.
When writing to dasd attributes (e.g. readonly), all values besides '1'
are handled like '0'.
Other sysfs-attributes like 'online' are checking for '1' and for '0'
and do not accept other
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Don't use small stacks when lockdep is used.
The lock dependency validator adds a bunch of extra stack frames to
the stack, which can cause stack overflows. Especially seen on 31 bit
where the small stack is only 4k.
Signed-off-by: Heiko Carstens <
From: Ursula Braun <[EMAIL PROTECTED]>
[S390] non-unique constant/macro identifiers.
Add some prefixes to constands defined in drivers/s390/net/qdio.h
and drivers/s390/lcs.h to make it possible to include the three
header files drivers/s390/net/qeth.h, drivers/s390/net/qdio.h and
drivers/net/s390
From: Horst Hummel <[EMAIL PROTECTED]>
[S390] Enhanced handling of failed termination requests.
In case a request timed out and termination did not work, the console was
flooded with retry messages (every 1/10s). Now we use a 5s delay per retry and
generate a more precise message.
Signed-off-by:
From: Michael Holzheu <[EMAIL PROTECTED]>
[S390] No panic for failed reboot
If reboot fails (e.g. because wrong devno has been specified by the user),
we should just stop all cpus, but should not trigger a kernel panic.
Signed-off-by: Michael Holzheu <[EMAIL PROTECTED]>
Signed-off-by: Martin Sch
From: Michael Holzheu <[EMAIL PROTECTED]>
[S390] Add ipl/reipl loadparm attribute.
If multiple kernel images are installed on one DASD, the loadparm can be used
to select the boot configuration. This patch introduces the following two new
sysfs attributes:
/sys/firmware/ipl/loadparm: shows loadp
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Use diag260 for memory size detection.
Avoid the tprot loop if diag260 works and reports that there are no
holes in memory. The tprot instruction can lead to a significant delay
in the ipl process if the virtual guest has a lot of memory and the
hos
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] termio <-> termios conversion error handling.
Get rid of our own user_termio_to_kernel_termios() and
kernel_termios_to_user_termio() macros which didn't check for errors
on user space accesses. Instead use the generic functions which
handle this pro
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Convert extmem spin_lock into a mutex.
There's no need to have a spin_lock here, but need sleepable context
for vmem_map. Therefore convert the spin_lock into a mutex.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidef
From: Christian Borntraeger <[EMAIL PROTECTED]>
[S390] set KBUILD_IMAGE.
Set KBUILD_IMAGE to a sane value. This enables "make rpm"
Signed-off-by: Christian Borntraeger <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
arch/s390/Makefile |3 +++
1 files changed,
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Add __must_check to uaccess functions.
Follow other architectures and add __must_check to uaccess functions.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
include/asm-s390/uaccess.h |
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] lockdep: show held locks when showing a stackdump
Follow i386/x86_64:
lockdep can be used to print held locks when printing a
backtrace. This can be useful when debugging things like
'scheduling while atomic' asserts.
Signed-off-by: Heiko Carstens
From: Cornelia Huck <[EMAIL PROTECTED]>
[S390] cio: Use device_reprobe() instead of bus_rescan_devices().
In io_subchannel_register(), it is better to just reprobe the current
device if it hasn't a driver yet than to rescan the whole bus.
Signed-off-by: Cornelia Huck <[EMAIL PROTECTED]>
Signed-o
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] cpcmd <-> __cpcmd calling issues
In case of reipl cpcmd gets called when all other cpus are not running
anymore. To prevent deadlocks change __cpcmd so that it doesn't take
any locks and call cpcmd or __cpcmd, whatever is correct in the current
cont
From: Ralph Wuerthner <[EMAIL PROTECTED]>
[S390] update interface notes in zcrypt.h
Signed-off-by: Ralph Wuerthner <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
include/asm-s390/zcrypt.h | 91 --
1 files changed, 41 i
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Memory detection fixes.
VMALLOC_END on 31bit should be 0x800UL instead of 0x7fffL.
The page mask which is used to make sure memory_end is on 4MB/2MB
boundary is wrong and not needed. Therefore remove it.
Make sure a vmalloc area does also ex
From: Gerald Schaefer <[EMAIL PROTECTED]>
[S390] Add dynamic size check for usercopy functions.
Use a wrapper for copy_to/from_user to chose the best usercopy method.
The mvcos instruction is better for sizes greater than 256 bytes, if
mvcos is not available a page table walk is better for sizes
From: Cornelia Huck <[EMAIL PROTECTED]>
[S390] cio: Retry internal operations after vary off.
If I/O was running on a just varied off chpid, it will be terminated.
If this was a common I/O layer internal I/O, it needs to be retried.
Signed-off-by: Cornelia Huck <[EMAIL PROTECTED]>
Signed-off-by:
From: Nicolas Kaiser <[EMAIL PROTECTED]>
[S390] Some documentation typos.
Signed-off-by: Nicolas Kaiser <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
Documentation/s390/CommonIO |4 +--
Documentation/s390/Debugging390.txt | 38
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] extmem unbalanced spin_lock.
segment save will exit with a lock held if the passed segment doesn't
exist. Any subsequent call to segment_save will lead to a deadlock.
Fix this and give up the lock before returning.
Signed-off-by: Heiko Carstens <[E
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Misaligned wait PSW at memory detection.
If the memory detection code would ever reach the point where it would
load the wait psw, it would generate a specification exception and the
system would crash at ipl time. This is because of a misaligned wa
From: Christian Borntraeger <[EMAIL PROTECTED]>
[S390] remove salipl memory detection.
The SALIPL entry point has an needless memory detection routine as we
later check the memory size again. The SALIPL code also uses diagnose
0x060 if we are running under VM, but this diagnose is not compatible
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] cpu shutdown rework
Let one master cpu kill all other cpus instead of sending an external
interrupt to all other cpus so they can kill themselves.
Simplifies reipl/shutdown functions a lot.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-o
From: Martin Schwidefsky <[EMAIL PROTECTED]>
[S390] 3215 device locking.
Remove lock pointer from 3215 device structure. Use get_ccwdev_lock
for each use of the lock in the ccw-device structure.
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
drivers/s390/char/con3215.c | 50 ++
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] pfault code cleanup.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
arch/s390/kernel/machine_kexec.c |8 +---
arch/s390/kernel/smp.c | 13 +++--
arch/s390/ker
From: Cornelia Huck <[EMAIL PROTECTED]>
[S390] cio: Use path verification for last path gone after vary off.
If the last path to a device is gone after a chpid has been varied
off, putting it on the slow queue doesn't prevent a device driver
from still attempting to use it (it may stay on the slo
From: Cornelia Huck <[EMAIL PROTECTED]>
[S390] Add MODALIAS= to the uevent for the ap bus.
Signed-off-by: Cornelia Huck <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
drivers/s390/crypto/ap_bus.c | 10 +-
1 files changed, 9 insertions(+), 1 deletion(-)
From: Horst Hummel <[EMAIL PROTECTED]>
[S390] return 'count' for successful execution of dasd_eer_enable.
Currently the return value of 'dasd_eer_enable' is returned - even if the
function returned '0'. Now return 'count' for successful execution.
Signed-off-by: Horst Hummel <[EMAIL PROTECTED]>
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Remove unused GENERIC_BUST_SPINLOCK from Kconfig.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
arch/s390/Kconfig |3 ---
1 files changed, 3 deletions(-)
diff -urpN linux-2.6/arch/
From: Michael Holzheu <[EMAIL PROTECTED]>
[S390] Use diag instead of ccw reipl.
Since the diag 308 reipl method is superior to the ccw method, we should
use it whenever it is possible. We can do that, if the user has not
specified a new reipl ccw device and the system has been ipled from
a ccw de
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Bad kexec control page allocation.
KEXEC_CONTROL_MEMORY_LIMIT is an unsigned long value and therefore
should be defined as one. Otherwise the kexec control page can be
allocated above 2GB which will cause a specification exception on the
sam31 instr
From: Heiko Carstens <[EMAIL PROTECTED]>
[S390] Cleanup memory_chunk array usage.
Need this at yet another file and don't want to add yet another
extern...
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Martin Schwidefsky <[EMAIL PROTECTED]>
---
arch/s390/kernel/head31.S |
On Mon, Dec 04, 2006 at 02:40:05PM +, Mel Gorman wrote:
> On (02/12/06 13:15), Nick Piggin didst pronounce:
> > Hi,
> >
> > While working in this area, I noticed a few things we do that may not
> > have a positive payoff under the most common conditions. Untested yet,
> > and probably needs a
Aucoin wrote:
From: Nick Piggin [mailto:[EMAIL PROTECTED]
We had customers see similar incorrect OOM problems, so I sent in some
patches merged after 2.6.16. Can you upgrade to latest kernel? (otherwise
I guess backporting could be an option for you).
I will raise the question of moving the ke
On Mon, Dec 04, 2006 at 03:56:24PM +0100, Martin Schwidefsky wrote:
...
> diff -urpN linux-2.6/include/asm-s390/cio.h
> linux-2.6-patched/include/asm-s390/cio.h
> --- linux-2.6/include/asm-s390/cio.h 2006-12-04 14:50:48.0 +0100
> +++ linux-2.6-patched/include/asm-s390/cio.h 2006-12-04 1
On Sun, Dec 03, 2006 at 11:56:40AM +0100, Geert Uytterhoeven wrote:
> On Sat, 2 Dec 2006, Russell King wrote:
> > On Sat, Dec 02, 2006 at 06:55:39PM +0100, Adrian Bunk wrote:
> > > This patch makes the contents of the userspace asm/setup.h header
> > > consistent on all architectures:
> > > - expo
On Mon, 2006-12-04 at 13:51 +0200, Janne Karhunen wrote:
> On 12/2/06, William Estrada <[EMAIL PROTECTED]> wrote:
> > Hi guys,
> >
> > I have been trying to make FC5's kernel do a boot
> > with an NFS root file system. I see the support is in the
> > kernel(?).
>
> Is this really properly possi
On Mon, 2006-12-04 at 08:35 +0100, Andreas Jellinghaus wrote:
> or I can send you the kernel patch as file and the xen hypervisor:
> -rw-r--r-- 1 root root 244694 2006-04-19 12:14 /boot/xen-3.0.2-2.gz
> -rw-r--r-- 1 root root 604942 2006-04-19 12:14
> /usr/src/kernel-patches/diffs/xen/linux-2.6.
Herbert Xu wrote:
Congestion control is always appropriate in a shared network. Please
note that congestion control does not conflict with the objectives of
UDP. For UDP, congestion control can simply mean dropping packets at
the source. DCCP is a good replacement for UDP that has congestion
c
Arnd Bergmann wrote:
On Friday 01 December 2006 21:01, Maynard Johnson wrote:
+static void notify_spus_active(void)
+{
+ int node;
+ for (node = 0; node < MAX_NUMNODES; node++) {
+ struct spu *spu;
+ mutex_lock(&spu_prio->active_mutex[node]);
+
On Mon, 2006-12-04 at 10:23 -0500, Josef Sipek wrote:
> > diff -urpN linux-2.6/include/asm-s390/cio.h
> > linux-2.6-patched/include/asm-s390/cio.h
> > --- linux-2.6/include/asm-s390/cio.h2006-12-04 14:50:48.0
> > +0100
> > +++ linux-2.6-patched/include/asm-s390/cio.h2006-1
Raz Ben-Jehuda(caro) wrote:
Who returns EGAIN to whom ? I am not sure i understand what you mean
here.
If the queue is full then io_sumbit() should return EAGAIN or some other
error to indicate that the queue is full, but right now it just blocks
instead.
-
To unsubscribe from this list:
* Sergei Shtylyov <[EMAIL PROTECTED]> wrote:
> >i'd suggest to redo it - but please keep it simple and clean. Those
> >dozens of casts to u64 are quite ugly.
>
>Alas, there's *nothing* I can do about it with 32-bit cycles_t.
> [...]
there's *always* a way to do such things more cleanly -
On Mon, 4 Dec 2006, Matthew Wilcox wrote:
>
> The sym2 driver would like to hash a u32 value, and it could just
> call hash_long() and rely on integer promotion on 64-bit machines, but
> that seems a little wasteful.
Hmm. It would appear that the file now needs
, no?
I do think it's a good ch
I don't have access to my ds1337 hardware right now (working away from
my office). Is someone else able to test this?
--
James Chapman
Katalix Systems Ltd
http://www.katalix.com
Catalysts for your Embedded Linux software development
David Brownell wrote:
On Saturday 02 December 2006 5:55 am,
In sound/isa/cmi8330.c, the ENABLE_SB_MIXER code is currently never
used.
What's the story behind this?
Should ENABLE_SB_MIXER be enabled?
Or the code be removed?
cu
Adrian
--
"Is there not promise of rain?" Ling Tan asked suddenly out
of the darkness. There had been need of ra
I think that I am seeing two seperate issues here that are getting mixed up.
1. while doing the tar + patch the system is choosing to use memory for
caching the tar (pushing program data out to cache).
2. after the tar has completed the data remins in the cache.
the answer for #2 is the one t
Russell King <[EMAIL PROTECTED]> wrote:
> I assume you wanted to delete "data" ?
Yes.
> Your premise is two timer_lists which use one common handler.
>
> struct foo {
> struct timer_list timer1;
> strucr timer_list timer2;
> };
That's not what I was thin
Nick Piggin <[EMAIL PROTECTED]> wrote:
> Move to an architecture independent rwsem implementation, using the
> better of the two rwsem implementations (ie. the one which doesn't
> take a spinlock to take an uncontested rwsem) as a basis. This gives
> us a single rwsem design instead of two.
Sigh.
On Mon, Dec 04, 2006 at 01:03:29PM +, David Howells wrote:
> Russell King <[EMAIL PROTECTED]> wrote:
> > I assume you wanted to delete "data" ?
>
> Yes.
>
> > Your premise is two timer_lists which use one common handler.
> >
> > struct foo {
> > struct timer_list timer1;
> >
Russell King <[EMAIL PROTECTED]> wrote:
> > You've now got four copies of struct timer_list, but only one handler.
>
> actioned.>
I was merely correcting your assumption of what premise I was thinking of.
You thought of something unnecessarily complex. I said nothing about what I
thought of yo
On Mon, Dec 04, 2006 at 02:17:21PM +, David Howells wrote:
> Russell King <[EMAIL PROTECTED]> wrote:
>
> > > You've now got four copies of struct timer_list, but only one handler.
> >
> > > actioned.>
>
> I was merely correcting your assumption of what premise I was thinking of.
> You thoug
On Mon, Dec 04, 2006 at 12:55:16PM +, David Howells wrote:
> Nick Piggin <[EMAIL PROTECTED]> wrote:
>
> > Move to an architecture independent rwsem implementation, using the
> > better of the two rwsem implementations (ie. the one which doesn't
> > take a spinlock to take an uncontested rwsem)
On Mon, Dec 04, 2006 at 07:38:00AM +0100, Oliver Neukum wrote:
> Am Montag, 4. Dezember 2006 05:43 schrieb Maneesh Soni:
> > On Fri, Dec 01, 2006 at 11:43:06PM +0100, Oliver Neukum wrote:
> > > Hi,
> > >
> > > Alan Stern has discovered a race in sysfs, whereby driver callbacks could
> > > be
> > >
Am Montag, 4. Dezember 2006 14:04 schrieb Maneesh Soni:
> > > Hi Oliver,
> > >
> > > Thanks for the explaining the patch but some description about the race
> > > would also help here. At the least the callpath to the race would be
> > > useful.
> > >
> > >
> > > Thanks
> > > Maneesh
> >
> > We h
Resend the patch.
The content of memory map io of BAR5 have been change from MCP65 then sata_nv
can't work fine on the platform based on MCP65 and MCP67,so move their IDs from
sata_nv.c to ahci.c.
Please check attachment for new patch,thanks.
Signed-off-by: Peer Chen <[EMAIL PROTECTED]>
-O
his patch adds VT6421 PATA support to sata_via.
If we don't want to enable PATA support, nothing is changed in sata_via
driver:
VIA SATA support
[ ] VT6421 PATA support (HIGHLY EXPERIMENTAL)
The patch is based on the patch from VIA, it applies to 2.6.19.
--
Tomasz Chmielewski
http:
On Mon, 04 Dec 2006 15:26:02 +0100
Tomasz Chmielewski <[EMAIL PROTECTED]> wrote:
> his patch adds VT6421 PATA support to sata_via.
NAK
This shouldn't be compile time
It doesn't handle MWDMA properly
It uses the old error/reset code
-
To unsubscribe from this list: send the line "unsubscribe lin
Wants testing... so test and report
Signed-off-by: Alan Cox <[EMAIL PROTECTED]>
--- linux.vanilla-2.6.19-rc6-mm1/drivers/ata/sata_via.c 2006-11-24
13:58:05.0 +
+++ linux-2.6.19-rc6-mm1/drivers/ata/sata_via.c 2006-12-04 14:57:34.719099648
+
@@ -59,11 +59,14 @@
SATA_INT_GA
ext Pierre Ossman wrote:
Anderson Briglia wrote:
Hi Pierre,
ext Pierre Ossman wrote:
Patch looks ok. But I never got an answer what the difference between
"change" and "assign" is.
You're right, the command is the same, but the difference is the
password's
length and password itself sent to t
Aucoin wrote:
The definition of perfectly good here may be up for debate or
someone can explain it to me. This perfectly good data was
cached under the tar yet hours after the tar has completed the
pages are still cached.
If nothing else has asked for that memory since the tar, there is no
re
On Mon, Dec 04, 2006 at 07:46:51AM -0800, Linus Torvalds wrote:
> On Mon, 4 Dec 2006, Matthew Wilcox wrote:
> >
> > The sym2 driver would like to hash a u32 value, and it could just
> > call hash_long() and rely on integer promotion on 64-bit machines, but
> > that seems a little wasteful.
>
> Hmm
On Fri, Dec 01, 2006 at 10:23:51AM -0500, Trond Myklebust wrote:
> On Fri, 2006-12-01 at 14:38 +, Sven Geggus wrote:
> > Hello,
> >
> > while trying to boot Kernel 2.6.19 (vanilla+unionfs) I get the following
> > NULL pointer dereferences:
>
> Are you able to reproduce this without unionfs?
ext Pierre Ossman wrote:
Anderson Briglia wrote:
This patch is needed only for lock/unlock commands. So, it's necessary to
make MMC omap works when using that feature. It's not a generic patch.
But I can take off this one from the series and send after (if) the
series
is integrated.
The patch
>Would it be possible to add printks for the EIOs? We want to know
>when dwarf2 is corrupted.
Here's a patch to do this and some more (applies on firstfloor tree, but
probably not on plain 2.6.19).
Signed-off-by: Jan Beulich <[EMAIL PROTECTED]>
--- linux-2.6.19-ff/kernel/unwind.c 2006-12-04
Now that binutils' ld is able to properly populate .eh_frame_hdr in the
Linux kernel case, here's a patch to add some functionality to the Dwarf2
unwinder to actually be able to make use of this (applies on firstfloor
tree with the previously sent patch to add debug output, but not on plain
2.6.19)
Hello.
Ingo Molnar wrote:
i'd suggest to redo it - but please keep it simple and clean. Those
dozens of casts to u64 are quite ugly.
Alas, there's *nothing* I can do about it with 32-bit cycles_t.
[...]
there's *always* a way to do such things more cleanly - such as the
patch below. Cou
The .eh_frame section contents is never written to, so it can as well
benefit from CONFIG_DEBUG_RODATA.
Diff-ed against firstfloor tree.
Signed-off-by: Jan Beulich <[EMAIL PROTECTED]>
Index: 2.6.19-ff-unwind-debug-msg/arch/i386/kernel/vmlinux.lds.S
===
* Sergei Shtylyov <[EMAIL PROTECTED]> wrote:
> > /* check for buggy clocks, handling wrap for 32-bit clocks */
> >-if (TYPE_EQUAL(cycles_t, unsigned long)) {
> >+if (TYPE_EQUAL(cycle_t, unsigned long)) {
> > if (time_after((unsigned long)T1, (unsigned long)T2))
> >
io_submit_one assigns ki_left = ki_nbytes = iocb->aio_nbytes, then
calls down to aio_setup_iocb, then to aio_setup_single_vector. In there,
ki_nbytes is reassigned to the same value it got two call stack above it.
There is no need to do so.
Signed-off-by: Ken Chen <[EMAIL PROTECTED]>
diff -Nurp
On Mon, 2006-12-04 at 16:27 +, Jan Beulich wrote:
> Kallsyms data is never written to, so it can as well benefit from
> CONFIG_DEBUG_RODATA.
>
> Signed-off-by: Jan Beulich <[EMAIL PROTECTED]>
Acked-by: Arjan van de Ven <[EMAIL PROTECTED]>
-
To unsubscribe from this list: send the line "unsu
Kallsyms data is never written to, so it can as well benefit from
CONFIG_DEBUG_RODATA.
Signed-off-by: Jan Beulich <[EMAIL PROTECTED]>
--- linux-2.6.19/kernel/kallsyms.c 2006-11-29 22:57:37.0 +0100
+++ 2.6.19-kallsyms-rodata/kernel/kallsyms.c2006-12-04 15:20:16.0
+0100
@@
Several drivers have bogus ioctl code that tries unneccessarily to
override the standard processing. In the three cases here the actual code
is not only wrong but also not required as they implement the proper
set_termios method as well.
Remove the junk.
Signed-off-by: Alan Cox <[EMAIL PROTECTED]
The access_ok() and negative length check on each iov segment in function
generic_file_aio_read/write are redundant. They are all already checked
before calling down to these low level generic functions.
Vector I/O (both sync and async) are checked via rw_copy_check_uvector().
For single segment
On Mon, 4 Dec 2006, Matthew Wilcox wrote:
>
> You're right. It hasn't actually caused any problems, presumably due to
> linux/types.h being included absolutely everywhere already, but it's
> clearly the Right Thing To Do. Updated patch:
I'm still not happy.
The return value of "hash_u64()" h
Signed-off-by: Alan Cox <[EMAIL PROTECTED]>
diff -u --new-file --recursive --exclude-from /usr/src/exclude
linux.vanilla-2.6.19-rc6-mm1/drivers/ata/pata_via.c
linux-2.6.19-rc6-mm1/drivers/ata/pata_via.c
--- linux.vanilla-2.6.19-rc6-mm1/drivers/ata/pata_via.c 2006-11-24
13:58:28.0 +
Switch to pci_get_bus_and_slot because some x86 systems seem to be
handing us a device with dev->bus = NULL. Also don't apply the isa fixup
to revision C6 and later of the chip.
Really we need to work out wtf is handing us pdev->bus = NULL, but firstly
and more importantly we need the drivers wor
On Sun, Dec 03, 2006 at 12:25:17AM +0300, Oleg Nesterov wrote:
> On top of rcu-add-a-prefetch-in-rcu_do_batch.patch
>
> rcu_do_batch:
>
> struct rcu_head *next, *list;
>
> while (list) {
> next = list->next; <-- [1]
> list->func(list);
>
When some objects are allocated by one CPU but freed by another CPU we can
consume lot of cycles doing divides in obj_to_index().
(Typical load on a dual processor machine where network interrupts are handled
by one particular CPU (allocating skbufs), and the other CPU is running the
applicatio
On Mon, 4 Dec 2006, Eric Dumazet wrote:
> Doing some math, we can use a reciprocal multiplication instead of a divide.
Could you generalize the reciprocal thingy so that the division
can be used from other parts of the kernel as well? It would be useful to
separately get some cycle counts on a
Hi,
I have modified the listio support patch by Sebastien to provide
the listio support through a new system call.
long lio_submit(aio_context_t ctx_id, int mode, long nr,
struct iocb __user * __user *iocbpp, struct sigevent __user *event)
More details about the system call appears withi
On Sun, 3 Dec 2006, Linus Torvalds wrote:
> Wouldn't it be much nicer to just lower the dirty-page limit?
>
> echo 1 > /proc/sys/vm/dirty_background_ratio
> echo 2 > /proc/sys/vm/dirty_ratio
Dirty ratio cannot be set to less than 5%. See
mm/page-writeback.c:get_dirty_limits().
> or
On Mon, Dec 04, 2006 at 01:21:29PM +0100, bert hubert wrote:
> On Sun, Dec 03, 2006 at 05:53:23PM -0800, Bill Huey wrote:
>
> > [8264, 996648, 0] {inode_init_once, fs/inode.c, 196}
> > [8552, 996648, 0] {inode_init_once, fs/inode.c, 193}
>
> Impressive, Bill!
>
> How tightly
Hello.
Ingo Molnar wrote:
/* check for buggy clocks, handling wrap for 32-bit clocks */
- if (TYPE_EQUAL(cycles_t, unsigned long)) {
+ if (TYPE_EQUAL(cycle_t, unsigned long)) {
if (time_after((unsigned long)T1, (unsigned long)T2))
prin
Aucoin <[EMAIL PROTECTED]> wrote:
[...]
> The definition of perfectly good here may be up for debate or
> someone can explain it to me. This perfectly good data was
> cached under the tar yet hours after the tar has completed the
> pages are still cached.
That means that there isn't a need for t
On Wed, Nov 29, 2006 at 11:32:34AM +0100, Sébastien Dugué wrote:
>
>
> +/***
> + * good_sigevent - check and get target task from a sigevent.
> + * @event: the sigevent to be checked
> + *
> + * This function must be called with tasklist_lock held for reading.
> + */
> +struct task_struct * good
On Monday 04 December 2006 17:29, Trond Myklebust wrote:
> > > I have been trying to make FC5's kernel do a boot
> > > with an NFS root file system. I see the support is in the
> > > kernel(?).
> >
> > Is this really properly possible (with read/write access and
> > locking in place)? AFAIK NFS
Alan,
On Mon, 4 Dec 2006 13:14:09 +, Alan wrote:
> On Mon, 4 Dec 2006 13:51:37 +0100 Jean Delvare wrote:
> > Any progress on this? I'd like to see this patch in -mm so that it gets
> > wider testing. One thing that needs to be fixed is that the Asus SMBus
> > quirks are currently ifdef'd out w
From: Randy Dunlap <[EMAIL PROTECTED]>
CC arch/i386/boot/compressed/misc.o
arch/i386/boot/compressed/misc.c:120: error: static declaration of
'xquad_portio' follows non-static declaration
include/asm/io.h:275: error: previous declaration of 'xquad_portio' was here
make[2]: *** [arch/i386/b
> From: David Lang [mailto:[EMAIL PROTECTED]
> I think that I am seeing two seperate issues here that are getting mixed
> up.
Fair enough.
> however the real problem that Aucoin is running into is patching process
> (tar, etc) kicks off the system is choosing to use it's
First name Louis, yes bu
Hi Tony, all,
On Thu, 10 Aug 2006 16:19:26 +0300, Tony Lindgren wrote:
> * Jean Delvare <[EMAIL PROTECTED]> [060810 11:30]:
> > I've now taken Komal's patch (#4). Here is a proposed patch which brings
> > the prescaler computation formula in line with your comment and table
> > above. It could be
> From: Horst H. von Brand [mailto:[EMAIL PROTECTED]
> That means that there isn't a need for that memory at all (and so they
In the current isolated non-production, not actually bearing a load test
case yes. But if I can't get it to not swap on an idle system I have no hope
of avoiding OOM on a
Hi.
Months ago two patches were posted by Ingo Molnar that removed
the 'volatile' keyword from the 'raw_spinlock_t' and 'raw_rwlock_t'
structures in the i386 and x86-64 ports. The link below takes you to
the initial (?) posting of the patches:
http://marc.theaimsgroup.com/?l=linux-kernel&m=115217
Hello.
Sergei Shtylyov wrote:
Why is cycles_t 32-bits on some
of the arches to begin with?
I guess this was done for speed reasons.
The whole 64-bit timebase can't be rad atomically on PPC32. ARM probably
has stricter limitations...
Ingo
WBR, Sergei
-
To unsubscribe fr
On Mon, 04 Dec 2006 14:07:22 -0300
"Horst H. von Brand" <[EMAIL PROTECTED]> wrote:
> Please explain again:
>
> - What you are doing, step by step
That 2GB machine apparently has a 1.6GB shm segment which is mlocked. That will
cause the VM to do one heck of a lot of pointless scanning and could,
On Mon, 4 Dec 2006, Andrew Morton wrote:
> but that's rather dumb. Better would be to remove mlocked pages from the
> LRU.
Could we generalize the removal of sections of a zone from the LRU? I
believe this would help various buffer allocation schemes. We have some
issues with heavy LRU scans i
101 - 200 of 423 matches
Mail list logo
