On Mon, Jan 24, 2005 at 07:58:03PM -0500, Brian Gerst wrote:
> Matthew Wilcox wrote:
> >__get_free_pages() calls alloc_pages, finds the page_address() and
> >throws away the struct page *. Slab then calls virt_to_page to get it
> >back again. Much more efficient for sla
On Tue, Jan 25, 2005 at 05:04:21PM +0530, Amit Gud wrote:
> Unify the spinlock initialization as far as possible.
>
> Do consider applying.
Actually, 'handler' and 'lock' are initialised for you (see
kernel/irq/handle.c) so I think those two lines can just be deleted.
'action' is also initialised
On Tue, Jan 25, 2005 at 12:22:16AM +1100, Keith Owens wrote:
> Unfortunately marking jiffies and similar small but high usage
> variables as section .sbss or .sdata requires changes to common code.
> It might be worth doing, but the change would have to be structured so
> it worked on all architect
On Tue, Nov 13, 2007 at 09:01:29AM -0800, Greg KH wrote:
> I'm still not sold on this idea at all. I'm really betting that there
> is a lot of incorrect acpi slot information floating around in machines
> and odd things will show up in these slot entries.
Is that the end of the world? Instead of
On Tue, Nov 13, 2007 at 12:50:08PM -0500, Mark Lord wrote:
> It's a 540MByte download over a slow link for everyone else.
Where do you get this number from?
$ du -sh .git/objects/pack/
249M.git/objects/pack/
$ du -sh .git/objects/
253M.git/objects/
ie about half what you claim.
--
Intel
On Tue, Nov 13, 2007 at 01:43:53PM -0500, Mark Lord wrote:
> Matthew Wilcox wrote:
> >ie about half what you claim.
> ..
>
> No, it's from earlier in this very thread:
>
> Adrian Bunk wrote:
> >git clone \
> >git://git.kernel.org/pub/scm/linux/kernel/git
On Tue, Nov 13, 2007 at 01:48:15PM -0600, Linas Vepstas wrote:
> On Mon, Nov 12, 2007 at 05:13:36PM -0700, Alex Chiang wrote:
> > + slot->name = kmalloc(8, GFP_KERNEL);
> > + sprintf(slot->name, "fake%d", count++);
>
> Please use snprintf to avoid buffer overruns!
Or, since kmalloc can return
On Tue, Nov 13, 2007 at 01:56:46PM -0600, Linas Vepstas wrote:
> On Mon, Nov 12, 2007 at 05:14:47PM -0700, Alex Chiang wrote:
> > +/* pci_slot represents a physical slot */
> > +struct pci_slot {
> > + struct pci_bus *bus;/* The bus this slot is on */
> > + struct pci_slot *next;
On Tue, Nov 13, 2007 at 10:51:22AM -0800, Greg KH wrote:
> Ok, again, I want to see the IBM people sign off on this, after testing
> on all of their machines, before I'll consider this, as I know the IBM
> acpi tables are "odd".
That seems a little higher standard than patches are normally held to
On Tue, Nov 13, 2007 at 03:41:21PM -0600, Linas Vepstas wrote:
> /sys/bus/pci/slots
> /sys/bus/pci/slots/control
> /sys/bus/pci/slots/control/remove_slot
> /sys/bus/pci/slots/control/add_slot
> /sys/bus/pci/slots/0001:00:02.0
> /sys/bus/pci/slots/0001:00:02.0/phy_location
Ugh. Almost two years ag
On Tue, Nov 13, 2007 at 02:56:05PM -0800, Greg KH wrote:
> Why not just use the code in the linux firmware kit that does this
> already today from userspace (thanks to Kristen for pointing this out to
> me on irc.)?
So then we have something that works on ACPI-based machines. Who will
add support
On Tue, Nov 13, 2007 at 03:33:14PM -0800, Kristen Carlson Accardi wrote:
> As far as being able to retrieve the slot number (which it seemed from
> the HP manageablity application perspective is the goal here), that
> information is available from userspace as well for at least standard PCI
> and p
On Wed, Nov 14, 2007 at 12:46:20AM -0700, Denys Vlasenko wrote:
> Finally they replied and asked to rediff it against their
> git tree. I did that and sent patches back. No reply since then.
>
> And mind you, the patch is not trying to do anything
> complex, it mostly moves code around, removes 'i
On Wed, Nov 14, 2007 at 02:07:56AM +0100, Andi Kleen wrote:
> It's not only complexity. Each new sysfs entry costs memory.
> Memory is not free. There should be always a good reason for those.
It's not a lot of memory; it's one directory and a couple of files for
each PCI slot in the system. Even
On Wed, Nov 14, 2007 at 03:53:15PM +0100, Ingo Molnar wrote:
> > By the way: Reverting commit 6f5391c283d7fdcf24bf40786ea79061919d1e1d
> > makes the same cd medium readable again on v2.6.24-rc2-409-g9418d5d.
>
> nice - that commit should then be reverted.
We're investigating; see bugzilla 9370.
On Wed, Nov 14, 2007 at 03:35:33PM +0100, Andi Kleen wrote:
> It becomes much more when someone does a find /sys. dentries are
> expensive. They eventually can get pruned again, but it's still
> costly to do that.
Again, if this is a big concern for you, there are better places to look
at for sav
On Wed, Nov 14, 2007 at 04:08:01PM +0100, Andi Kleen wrote:
> Whoever is proposing a feature has the burden to justify that
> its usefulness is larger than the overhead/cost it adds.
>
> Doesn't seem to be the case with this one so far.
Huh? There are half a dozen people who think it does, and h
On Wed, Nov 14, 2007 at 12:37:29PM -0700, Alex Chiang wrote:
> Register one slot per slot, rather than one slot per function.
> Change the name of the slot to fake%d instead of the pci address.
> +#define SLOT_NAME_SIZE KOBJ_NAME_LEN
Defined, then never used ... how about s/KOBJ_NAME_LEN/8/, then
On Mon, Nov 19, 2007 at 04:35:23PM +1100, Benjamin Herrenschmidt wrote:
> The other one I'm hitting now is that the SCSI layer nowadays embeds the
'nowadays'? It has always been so.
> sense_buffer inside the scsi_cmnd structure without any kind of
> alignment whatsoever. I've been hitting irregu
On Thu, Oct 25, 2007 at 10:07:17PM -0400, Jeff Garzik wrote:
> Is this warning of value to anybody but you?
Yup. It signals this driver isn't production quality yet.
> Is this information worth printing out on everyone else's kernel build?
I thnk that's worth noting to anyone who's enabled it.
On Fri, Oct 26, 2007 at 12:11:01PM +1000, Rusty Russell wrote:
> This just seems like more optimization and complexity that we need.
> Interfaces
> using vsnprintf don't seem like good candidates for optimization.
That's a fair point, but I'm optimising for fewer trips into the
slab(/slub/
On Sat, Oct 27, 2007 at 12:39:41AM +0400, Alexey Dobriyan wrote:
> > > - if ((task->state != TASK_STOPPED) && (task->state !=
> > > TASK_TRACED)) {
> > > + if (!is_task_stopped_or_traced(task)) {
> ^^
>
> I think this is horrible. Are y
On Sat, Oct 27, 2007 at 07:34:47PM +0300, Pekka Enberg wrote:
> On 10/27/07, Rusty Russell <[EMAIL PROTECTED]> wrote:
> > On Saturday 27 October 2007 21:47:09 Pekka Enberg wrote:
> > > Why don't we just return -ENOMEM here just like all other APIs in the
> > > kernel?
> > I think Willy did it becau
On Sun, Oct 28, 2007 at 04:51:42PM +0100, Adrian Bunk wrote:
> This patch removes obviously dead code.
> case LINKED_WRONG:
> WRITE_DATA (ABORT);
> linked_connected = 0;
>
On Sun, Oct 28, 2007 at 01:43:21PM -0400, J. Bruce Fields wrote:
> We currently attempt to return -EDEALK to blocking fcntl() file locking
> requests that would create a cycle in the graph of tasks waiting on
> locks.
>
> This is inefficient: in the general case it requires us determining
> whethe
On Sun, Oct 28, 2007 at 06:40:52PM +, Alan Cox wrote:
> NAK. This is an ABI change. It was also comprehensively rejected before
> because
>
> - EDEADLK behaviour is ABI
Not in any meaningful way.
> - EDEADLK behaviour is required by SuSv3
What SuSv3 actually says is:
If the system
On Sun, Oct 28, 2007 at 05:50:30PM -0400, Trond Myklebust wrote:
> > You can't fix the false EDEADLK detection without solving the halting
> > problem. Best of luck with that.
>
> I can see that it would be difficult to do efficiently, but basically,
> this boils down to finding a circular path i
On Sun, Oct 28, 2007 at 10:48:33PM +, Alan Cox wrote:
> > Bzzt. You get a false deadlock with multiple threads like so:
> >
> > Thread A of task B takes lock 1
> > Thread C of task D takes lock 2
> > Thread C of task D blocks on lock 1
> > Thread E of task B blocks on lock 2
>
> The spec and
On Sun, Oct 28, 2007 at 11:55:52PM +0100, Jiri Kosina wrote:
> On Sun, 28 Oct 2007, Matthew Wilcox wrote:
>
> > Bzzt. You get a false deadlock with multiple threads like so:
> > Thread A of task B takes lock 1
> > Thread C of task D takes lock 2
> > Thread C of task
On Sun, Oct 28, 2007 at 09:38:55PM +, Alan Cox wrote:
> > It doesn't require the system to detect it, only mandate what to return
> > if it does detect it.
>
> We should be detecting at least the obvious case.
What is the obvious case? A task that has never called clone()?
> If SYSV only sp
On Mon, Oct 29, 2007 at 10:03:16PM -0500, Kumar Gala wrote:
> The same is true on PPC32. Its a per platform thing. However, I'm
> not sure if we could hide it from the user. There are cases on the
> same HW platform that you want to run with just 32-bit phys (for
> performance).
Have you
On Tue, Oct 30, 2007 at 11:42:21AM +, Ralf Baechle wrote:
> On Tue, Oct 30, 2007 at 02:37:19AM -0600, Matthew Wilcox wrote:
> > On Mon, Oct 29, 2007 at 10:03:16PM -0500, Kumar Gala wrote:
> > > same HW platform that you want to run with just 32-bit phys (for
> > >
On Tue, Oct 30, 2007 at 11:54:22AM +0100, Richard Knutsson wrote:
> Convert to use the generic boolean.
> - u8 initialized;
> - u8 in_use; /* is the management node open*/
> + bool initialized:8;
> + bool in_use:8; /* is the management node open*/
Are you seri
On Tue, Oct 30, 2007 at 04:02:25PM +0100, Richard Knutsson wrote:
> Matthew Wilcox wrote:
> >On Tue, Oct 30, 2007 at 11:54:22AM +0100, Richard Knutsson wrote:
> >
> >>Convert to use the generic boolean.
> >>- u8 initialized;
> >>- u8 in_use;
On Tue, Oct 30, 2007 at 05:46:08PM +0100, Richard Knutsson wrote:
> I just don't see the reason why expressing a boolean as an integer. Some
> advantage?
This is C, not Java, or some other highly-typed language.
if (int) and if (ptr) are perfectly acceptable in C.
> (also helps us if someone doe
On Sun, Oct 14, 2007 at 04:00:10PM +0100, Denys Vlasenko wrote:
> Deinlines and moves big functions from .h to .c files.
> Adds prototypes for ahc_lookup_scb and ahd_lookup_scb to .h files.
Adds trailing whitespace.
.dotest/patch:216: * We also set the full residual flag which the
Adds trai
On Sun, Oct 14, 2007 at 04:02:15PM +0100, Denys Vlasenko wrote:
> Adds more consts
error: patch failed: drivers/scsi/aic7xxx/aic79xx.h:1328
error: drivers/scsi/aic7xxx/aic79xx.h: patch does not apply
error: patch failed: drivers/scsi/aic7xxx/aic7xxx.h:1145
error: drivers/scsi/aic7xxx/aic7xxx.h: pa
On Thu, Nov 01, 2007 at 03:23:28PM -0700, Randy Dunlap wrote:
> when attempting to apply to scsi-misc or linus-git or ... ?
scsi-misc
--
Intel are signing my paycheques ... these opinions are still mine
"Bill, look, we understand that you're interested in selling us this
operating system, but co
make this look like:
#else /* ADVANSYS_DEBUG */
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
diff --git a/scripts/Lindent b/scripts/Lindent
index 7d8d889..9468ec7 100755
--- a/scripts/Lindent
+++ b/scripts/Lindent
@@ -1,2 +1,2 @@
#!/bin/sh
-indent -npro -kr -i8 -ts8 -sob -l
On Fri, Jul 20, 2007 at 09:20:15AM -0700, Luck, Tony wrote:
> Which is better. But if we unconditionally set this CONFIG variable, then the
> code in fs/quota.c will have to read:
>
> #if defined(CONFIG_COMPAT) && defined(CONFIG_COMPAT_FOR_U64_ALIGNMENT)
>
> We can keep it simpler if the Kconfig
On Sun, Jul 15, 2007 at 10:53:45AM +0100, Simon Arlott wrote:
> Do people ever check what Lindent does?
[...]
> > -out1:
> > + out1:
>
> NAK: A perfectly valid non-indented label is now indented by 6 spaces.
I tracked down why indent does this. It's actually hard-coded to indent
by 2 fewer
On Sun, Jul 15, 2007 at 10:53:45AM +0100, Simon Arlott wrote:
> > - } else if (base_addr > 0x100) { /* Check a single specified location. */
> > + } else if (base_addr > 0x100) { /* Check a single specified location. */
>
> What is Lit doing here?! It's changed "{/*" to "{/*"...
>
> > - } e
On Fri, Jul 20, 2007 at 06:31:32PM +0100, Ralf Baechle wrote:
> On Fri, Jul 20, 2007 at 06:40:46PM +0200, Geert Uytterhoeven wrote:
> > +#include
>
> These days that should probably be .
Go and read the comments at the top of linux/irq.h.
And then report to Russell for your whipping.
--
"Bill,
On Fri, Jul 20, 2007 at 09:43:47PM +0400, Sergei Shtylyov wrote:
> Hello Christoph:
>
> >>>+#include
>
> >>These days that should probably be .
>
> >Not at all, linux/irq.h is something entirely different.
>
>Actually,
Not for enable/disable_irq. For request_irq, yes.
This is something
On Fri, Jul 20, 2007 at 06:36:02PM +0100, Simon Arlott wrote:
> On 20/07/07 18:07, Matthew Wilcox wrote:
> > sed -i -e 's/^\t* \(\w*:\)/ \1/' "$@"
> >
> > which will replace the leading tabs and spaces with one space.
>
> ... isn't the sp
On Fri, Jul 20, 2007 at 10:50:42PM +0200, Oleg Verych wrote:
> > Of course, we can't add this flag to Lindent until it's widely
> > circulating amongst the distributions. Perhaps we can add this to
> > Lindent in the meantime:
> >
> > sed -i -e 's/^\t* \(\w*:\)/ \1/' "$@"
> >
> > which will r
On Sat, Jul 21, 2007 at 07:11:01AM +0100, Simon Arlott wrote:
> Changing the code to fix a utility bug is madness. I think it's been
> fixed too...
I just downloaded the source from CVS and it hasn't:
--- test1.c 2007-07-21 09:49:02.0 -0400
+++ test2.c 2007-07-21 09:49:12.
On Sun, Jul 22, 2007 at 03:52:21PM +0200, Gabriel C wrote:
> drivers/scsi/sym53c416.c: In function 'sym53c416_detect':
> drivers/scsi/sym53c416.c:638: warning: the address of 'sym53c416' will always
> evaluate as 'true'
> drivers/scsi/sym53c416.c:644: warning: the address of 'sym53c416_1' will
>
I noticed that we only look at the first action in the chain when
determining whether to re-enable local interrupts during handle_IRQ_event.
But we don't try to exclude sharing interrupts with mixtures of
IRQF_DISABLED set and clear. I just tried to do that locally, and one
of my USB ports disapp
On Thu, Jul 26, 2007 at 08:46:47AM -0400, James Bottomley wrote:
> On Thu, 2007-07-26 at 05:21 -0400, Jeff Garzik wrote:
> > libsas should -not- require PCI, even though aic94xx does.
>
> Realistically, even for parisc, I can't see anyone producing a non-PCI
> SAS device (even though I'd like one)
On Wed, Aug 29, 2007 at 01:38:04PM +1000, Nick Piggin wrote:
> Hasn't the KS committee / TAB board vote rigging conspiracy theory been
> raised yet?
It's too easy. All you have to do is note the significant overlap
between the KS program committee and the TAB.
Program Committee
Jens Axboe, Or
I've long hated the non-killability of tasks accessing a dead
NFS server. Linus had an idea for fixing this way back in 2002:
http://www.ussg.iu.edu/hypermail/linux/kernel/0208.0/0167.html which
I've prototyped in this patch.
Splitting up TASK_* into separate bits is going to need a lot more
aud
On Thu, Aug 30, 2007 at 12:40:38PM +0200, [EMAIL PROTECTED] wrote:
> - diff_int = my_abs(rate_ext-rate);
> - diff_ext = my_abs(rate_int-rate);
> + diff_int = abs(rate_ext-rate);
> + diff_ext = abs(rate_int-rate);
Nothing to do with the patch, but is
On Thu, Aug 30, 2007 at 09:11:37AM -0400, Trond Myklebust wrote:
> > diff --git a/fs/read_write.c b/fs/read_write.c
> > index 507ddff..29e3f21 100644
> > --- a/fs/read_write.c
> > +++ b/fs/read_write.c
> > @@ -220,7 +220,7 @@ Einval:
> >
> > static void wait_on_retry_sync_kiocb(struct kiocb *ioc
On Thu, Aug 30, 2007 at 08:43:24PM +0200, Jan Engelhardt wrote:
> On Aug 28 2007 20:55, Denys Vlasenko wrote:
> >> Fwiw I do like your debloat patch a lot; it's just only half the
> >> equation ... if you also do the namespace fixes, I bet the driver
> >> debloats even more...
> >
> >Yes, I know, a
On Thu, Aug 30, 2007 at 03:46:56PM -0700, Jason Gaston wrote:
> This updated patch adds the Intel Tolapai LPC and SMBus Controller DID's.
> +#define PCI_DEVICE_ID_INTEL_Tolapai_00x5031
> +#define PCI_DEVICE_ID_INTEL_Tolapai_10x5032
NACK -- use all upper-case.
--
Intel are signing
On Fri, Aug 31, 2007 at 02:53:23PM +0100, Christoph Hellwig wrote:
> On Thu, Aug 30, 2007 at 10:55:49PM +0200, Jan Engelhardt wrote:
> > "those callers". There was _exactly one_ caller, and that was an out-of-tree
> > module. There were not any in-kernel callers before, and it did not generate
> >
On Fri, Aug 31, 2007 at 08:42:59AM -0700, Andrew Morton wrote:
> On Fri, 31 Aug 2007 08:17:48 -0700 (PDT) [EMAIL PROTECTED] wrote:
>
> > http://bugzilla.kernel.org/show_bug.cgi?id=8942
>
> Seems that DAC960 broke in 2.6.22 - three people are reporting this.
Interesting. Because nobody touched D
[Note: not tested. I have a temporary shortage of machines with 5 volt
PCI slots, and in any case, I think the DAC960 that I have is a V1 card]
Use PCI_DMA_* constants instead of own private definitions
Fall back to 32-bit DMA mask if a 64-bit one fails
Signed-off-by: Matthew Wilcox <[EM
Here's the second version of TASK_KILLABLE. A few changes since version 1:
- Don't split up TASK_INTERRUPTIBLE and TASK_UNINTERRUPTIBLE.
TASK_WAKESIGNAL and TASK_LOADAVG were pretty much equivalent, and since
we had to keep __TASK_{UN,}INTERRUPTIBLE anyway, splitting them made
little s
Replace the uses of __wake_up_locked with wake_up_locked
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs/eventpoll.c | 11 ---
1 files changed, 4 insertions(+), 7 deletions(-)
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 77b9953..72e4cb4 100644
--- a/fs/event
and associated infrastructure such as sync_page_killable and
fatal_signal_pending. Use lock_page_killable in do_generic_mapping_read()
to allow us to kill `cat' of a file on an NFS-mounted filesystem.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/pagem
Set TASK_WAKEKILL for TASK_STOPPED and TASK_TRACED, add TASK_KILLABLE and
use TASK_WAKEKILL in signal_wake_up()
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/sched.h | 22 ++
kernel/signal.c |8
2 files changed, 18 insertions(
Abstracting away direct uses of TASK_ flags allows us to change the
definitions of the task flags more easily.
Also restructure do_wait() a little
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
arch/ia64/kernel/perfmon.c |4 +-
fs/proc/array.c|9 +---
fs/proc/
Use TASK_KILLABLE to allow wait_on_retry_sync_kiocb to return -EINTR.
All callers then check the return value and break out of their loops.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs/read_write.c | 17 -
1 files changed, 12 insertions(+), 5 deletions(-)
On Sat, Sep 01, 2007 at 10:46:51PM -0400, Matthew Wilcox wrote:
> Abstracting away direct uses of TASK_ flags allows us to change the
> definitions of the task flags more easily.
> --- a/kernel/exit.c
> +++ b/kernel/exit.c
> @@ -259,7 +259,7 @@ static int has_stopped_jobs(s
On Sat, Sep 01, 2007 at 08:35:06PM -0700, Daniel Walker wrote:
> Does it take task->state or task ?
task. Clearly I didn't test on ia64. (There was an iteration where it
took task->state, and I guess I missed one). Thanks for pointing out
this oops, I'll fix it for round three.
--
Intel are s
On Sat, Sep 01, 2007 at 10:46:51PM -0400, Matthew Wilcox wrote:
> Abstracting away direct uses of TASK_ flags allows us to change the
> definitions of the task flags more easily.
>
> Also restructure do_wait() a little
>
> Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED
On Thu, Sep 06, 2007 at 01:35:08PM +0800, Shaohua Li wrote:
> PCIE 'Enable No Snoop' bit is set by default per PCIE spec, but OS
> assumes PCI DMA is snooped, which is legacy PCI device does. Enabling no
> snoop might cause potential DMA issues. This patch disables it, if a
> driver can work with n
On Wed, Aug 01, 2007 at 09:39:12PM +0800, Eugene Teo wrote:
> This patch fixes the following compile error:
>
> drivers/scsi/advansys.c: In function 'advansys_board_found':
> drivers/scsi/advansys.c:17781: error: implicit declaration of function
> 'to_pci_dev'
Or just remove the ifdefs around the
On Wed, Aug 01, 2007 at 04:27:15PM +0200, Gabriel C wrote:
> Matthew Wilcox wrote:
> > I'd be interested in seeing the results of the randconfig trials on the
> > driver with those 23 patches applied, but not particularly interested in
> > the intermediate result.
>
On Wed, Aug 01, 2007 at 05:17:15PM +0200, Michal Piotrowski wrote:
> -#include
> +#include
Why does this driver need stat.h at all?
--
"Bill, look, we understand that you're interested in selling us this
operating system, but compare it to ours. We can't possibly take such
a retrograde step."
On Tue, Jul 31, 2007 at 09:18:08AM -0500, James Bottomley wrote:
> The other comment is that power saving seems to be a property of the
> transport rather than the host. If you do it in the transport classes,
> then you can expose all the knobs the actual transport possesses (which
> is, unfortuna
On Wed, Aug 01, 2007 at 05:26:53PM -0700, Andrew Morton wrote:
> Why on earth is that using GFP_ATOMIC? This function later goes on to
> create procfs files and such things.
Seems fairly common in driver initialisation code. I removed three
instances of this in the advansys driver.
> y'know, we
On Sun, Aug 12, 2007 at 01:43:16AM +0200, Jesper Juhl wrote:
> Trying to build current Linus git tree (head at
> ac07860264bd2b18834d3fa3be47032115524cea) using the attached config
> file (generated by 'make randconfig') the build fails for me with :
>
> ...
> CC drivers/scsi/advansys.o
On Sat, Aug 11, 2007 at 04:04:54PM +0100, Jurij Smakov wrote:
> [Please keep me on CC, as I'm not on LKML.]
> I've recently got a Sun Blade 1000 box with a QLA2200 controller, and
> I'm bumping into exact same problem with 2.6.22:
Please try
http://marc.info/?l=linux-scsi&m=118289275414202
which
On Sun, Aug 12, 2007 at 11:39:41PM -0700, [EMAIL PROTECTED] wrote:
> Add file pattern to MAINTAINER entry
Uh, what the fuck? You're posting over *500* patches to the same file?
Just post one patch and people can respond to the bits of it that
concern them.
--
"Bill, look, we understand that you
On Tue, Aug 14, 2007 at 11:10:22AM +0200, Tore Anderson wrote:
> I gave it a spin, and got quite a few troubles that appears related to
> the lpfc driver. I don't know if these problems happened due to the
> recent update as the latest kernel I ran before was 2.6.20 (where I
> never saw probl
On Wed, Aug 15, 2007 at 10:08:17AM -0400, Igor A. Nesterov wrote:
> So would it be true to say that the fix for -EEXIST problem still has
> not found its way to mainline kernel? I've been hit by this problem
> after switching to Fedora 7, and currently running on Fedora
> 2.6.21-1.3228 kernel patch
The 32-bit version is more efficient (and apparently gives better hash
results than the 64-bit version), so users who are only hashing a 32-bit
quantity can now opt to use the 32-bit version explicitly, rather than
promoting to a long.
Signed-off-by: Matthew Wilcox <[EMAIL PROTEC
Set TASK_WAKEKILL for TASK_STOPPED and TASK_TRACED, add TASK_KILLABLE and
use TASK_WAKEKILL in signal_wake_up()
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/sched.h | 22 ++
kernel/signal.c |8
2 files changed, 18 insertions(
This series of patches introduces the facility to deliver only fatal
signals to tasks which are otherwise waiting uninterruptibly.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/m
Replace the uses of __wake_up_locked with wake_up_locked
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs/eventpoll.c | 11 ---
1 files changed, 4 insertions(+), 7 deletions(-)
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 77b9953..72e4cb4 100644
--- a/fs/event
Use TASK_KILLABLE to allow wait_on_retry_sync_kiocb to return -EINTR.
All callers then check the return value and break out of their loops.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs/read_write.c | 17 -
1 files changed, 12 insertions(+), 5 deletions(-)
and associated infrastructure such as sync_page_killable and
fatal_signal_pending. Use lock_page_killable in do_generic_mapping_read()
to allow us to kill `cat' of a file on an NFS-mounted filesystem.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/pagem
Abstracting away direct uses of TASK_ flags allows us to change the
definitions of the task flags more easily.
Also restructure do_wait() a little
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
arch/ia64/kernel/perfmon.c |4 +-
fs/proc/array.c|9 +---
fs/proc/
On Thu, Oct 18, 2007 at 08:56:23PM -0700, Arjan van de Ven wrote:
> Have you tested this patch with LOCKDEP enabled? eventpoll is... tricky
> in what it does with waitqueues and locks and some of this stuff is
> there, afaik, to deal with that. You're now changing this ... call me
> chicken :)
On Mon, Oct 22, 2007 at 09:10:34PM +0200, Arnd Bergmann wrote:
> On a related note, should we encourage the use of spin_lock() and
> spin_lock_irq() instead of spin_lock_irqsave() where possible?
spin_lock(), certainly. On PowerPC, I'm reliably informed it's fewer
instructions to save/restore tha
Introduce new __holds() macro to tell sparse it's OK to drop and then
reacquire a lock within a function. Use it in scsi_request_fn.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 207f1aa..5e0583a 100644
--- a/d
r for it.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/stringbuf.h | 85 +
lib/Makefile |2 +-
lib/stringbuf.c | 79 +
3 files changed, 165 insertions(+),
With asynchronous disk scanning, it's quite common for the partition
report to be interrupted by reports of other discs being discovered.
Use a new function, print_partition, to accumulate the partitions into
a stringbuf.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs
Get rid of the _cdebbuf structure that was used to accumulate strings
for a debug printk and use the stringbuf instead. Allocate the stringbuf
on the stack instead of with kmalloc. Return a char * to the callers
rather than a stringbuf.
Signed-off-by: Matthew Wilcox <[EMAIL PROTEC
On Mon, Oct 22, 2007 at 07:09:56PM -0700, Linus Torvalds wrote:
> Umm. This is why we write things like
>
> static void double_lock_balance(struct rq *this_rq, struct rq *busiest)
> __releases(this_rq->lock)
> __acquires(busiest->lock)
> __acquires(t
On Tue, Oct 23, 2007 at 05:11:16PM -0500, Matt Mackall wrote:
> You might want to consider growing the buffer by no less than a small
> constant factor like 1.3x. This will keep things that do short concats
> in a loop from degrading to O(n^2) performance due to realloc and
> memcpy.
I looked at s
On Tue, Oct 23, 2007 at 04:43:53PM -0700, Linus Torvalds wrote:
> On Tue, 23 Oct 2007, Matthew Wilcox wrote:
> > This is a generic string buffer which can also be used for non-printk
> > purposes. There is no sb_scanf implementation yet as I haven't identified
> > a u
On Tue, Oct 23, 2007 at 10:19:06PM -0400, Eric St-Laurent wrote:
> I don't know if copy-on-write semantics are really useful for current
> in-kernel uses, but I've coded and used a C++ string class like this in
> the past:
CoW isn't in the slightest bit helpful. The point of these is to
provide a
Replace the uses of __wake_up_locked with wake_up_locked
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs/eventpoll.c | 11 ---
1 files changed, 4 insertions(+), 7 deletions(-)
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 34f68f3..81c04ab 100644
--- a/fs/event
Set TASK_WAKEKILL for TASK_STOPPED and TASK_TRACED, add TASK_KILLABLE and
use TASK_WAKEKILL in signal_wake_up()
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/sched.h | 22 ++
kernel/signal.c |8
2 files changed, 18 insertions(
and associated infrastructure such as sync_page_killable and
fatal_signal_pending. Use lock_page_killable in do_generic_mapping_read()
to allow us to kill `cat' of a file on an NFS-mounted filesystem.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
include/linux/pagem
Use TASK_KILLABLE to allow wait_on_retry_sync_kiocb to return -EINTR.
All callers then check the return value and break out of their loops.
Signed-off-by: Matthew Wilcox <[EMAIL PROTECTED]>
---
fs/read_write.c | 17 -
1 files changed, 12 insertions(+), 5 deletions(-)
301 - 400 of 5365 matches
Mail list logo
