Hello,
On Thu, Mar 28, 2024 at 01:53:25PM +0100, Ulf Hansson wrote:
> At this point we have suggested to drivers to switch to use threaded
> irq handlers (and regular work queues if needed too). That said,
> what's the benefit of using the BH work queue?
BH workqueues should behave about the
Hello,
On Mon, Feb 05, 2024 at 03:24:17PM +0530, Tasmiya Nalatwad wrote:
> Greetings,
>
> [Linux-next] [6.8.0-rc2-next-20240130] [FC / XFS] Task hangs for infinite
> time while running bonnie test XFS filesystem
>
> Bisected the issue. Git bisect points to the below commit
> commit
On Tue, May 09, 2023 at 05:09:43PM +1000, Michael Ellerman wrote:
> Stephen Rothwell writes:
> > Hi all,
> >
> > Today's qemu test boot (powerpc pseries_le_defconfig) produced this
> > warning:
> >
> > [2.048588][T1] ipr: IBM Power RAID SCSI Device Driver version:
> > 2.6.4 (March 14,
Applied to wq/for-6.5-cleanup-ordered.
Thanks.
--
tejun
ays
reconsider later.
As there are follow-up workqueue core changes, I'd really appreciate if the
patch can be routed through the workqueue tree w/ your acks. Thanks.
Signed-off-by: Tejun Heo
Cc: Michael Ellerman
Cc: Nicholas Piggin
Cc: Christophe Leroy
Cc: Nathan Lynch
Cc: linuxppc-
On Mon, Jun 29, 2020 at 04:07:01PM +0530, Srikar Dronamraju wrote:
> Tejun Heo had modified shared_cpu_map_show to use scnprintf instead of
> cpumap_print during support for *pb[l] format.
> Refer commit 0c118b7bd09a ("powerpc: use %*pb[l] to print bitmaps including
> cpuma
Hello, Laurent.
On Thu, Aug 24, 2017 at 02:10:31PM +0200, Laurent Vivier wrote:
> > Yeah, it just needs to match up new cpus to the cpu ids assigned to
> > the right node.
>
> We are not able to assign the cpu ids to the right node before the CPU
> is present, because firmware doesn't provide
Hello, Michael.
On Wed, Aug 23, 2017 at 09:00:39PM +1000, Michael Ellerman wrote:
> > I don't think that's true. The CPU id used in kernel doesn't have to
> > match the physical one and arch code should be able to pre-map CPU IDs
> > to nodes and use the matching one when hotplugging CPUs. I'm
Hello, Michael.
On Tue, Aug 22, 2017 at 11:41:41AM +1000, Michael Ellerman wrote:
> > This is something powerpc needs to fix.
>
> There is no way for us to fix it.
I don't think that's true. The CPU id used in kernel doesn't have to
match the physical one and arch code should be able to
On Mon, Aug 21, 2017 at 03:49:51PM +0200, Laurent Vivier wrote:
> cpumask is the list of CPUs present when the queue is built.
> If a new CPU is hotplugged, this list is not updated,
> and when the scheduler asks for a CPU id, blk_mq_hctx_next_cpu()
> can return WORK_CPU_UNBOUND.
> And in this
On Mon, Aug 21, 2017 at 03:49:50PM +0200, Laurent Vivier wrote:
> In wq_numa_init() a list of NUMA nodes with their list of possible CPUs
> is built.
>
> Unfortunately, on powerpc, the Firmware is only able to provide the
> node of a CPU if the CPU is present. So, in our case (possible CPU)
> CPU
>From 610467270fb368584b74567edd21c8cc5104490f Mon Sep 17 00:00:00 2001
From: Tejun Heo <t...@kernel.org>
Date: Sat, 8 Jul 2017 07:17:02 -0400
Subsystem migration methods shouldn't be called for empty migrations.
cgroup_migrate_execute() implements this guarantee by bail
Hello, Abdul.
Thanks for the debug info. Can you please see whether the following
patch fixes the issue? If the problem is too difficult to reproduce
to confirm the fix by seeing whether it no longer triggers, please let
me know. We can instead apply a patch which triggers WARN on the
failing
Hello, Michael.
On Fri, Jun 30, 2017 at 11:08:22AM +1000, Michael Ellerman wrote:
> Tejun Heo <t...@kernel.org> writes:
>
> > Could be the same problem as the one reported in the following thread.
> >
> > http://lkml.kernel.org/r/1497266622.15415.39.ca...@abdul.in.
Hello,
Could be the same problem as the one reported in the following thread.
http://lkml.kernel.org/r/1497266622.15415.39.ca...@abdul.in.ibm.com
The root cause there is ppc arch code not setting up possible cpu <->
numa mapping during boot.
Thanks.
--
tejun
Hello, Abdul.
Sorry about the long delay.
On Mon, Jun 12, 2017 at 04:53:42PM +0530, Abdul Haleem wrote:
> linux-next kernel crashed while running CPU offline and online.
>
> Machine: Power 8 LPAR
> Kernel : 4.12.0-rc4-next-20170609
> gcc : version 5.2.1
> config: attached
> testcase: CPU off/on
On Thu, Jun 22, 2017 at 12:31:35AM +0100, Chris Wilson wrote:
> Quoting Tejun Heo (2017-06-13 14:58:49)
> > Cc'ing David Airlie.
> >
> > This is from drm driver calling in idr_replace() w/ a negative id.
> > Probably a silly bug in error handling path?
>
On Tue, May 30, 2017 at 01:24:06PM +0530, Abdul Haleem wrote:
> Hi,
>
> Test : stress-ng
> Machine : Power 8 Bare Metal
> Kernel : 4.12.0-rc3
> Config : attached
> gcc version: 4.8.5
>
>
> In file kernel/workqueue.c at line 2041
>
> /* ensure we're on the correct CPU */
>
Cc'ing David Airlie.
This is from drm driver calling in idr_replace() w/ a negative id.
Probably a silly bug in error handling path?
Thanks.
On Mon, Jun 12, 2017 at 08:10:54PM +0530, Abdul Haleem wrote:
> Hi,
>
> WARN_ON_ONCE is being called from idr_replace() function in file
> lib/idr.c at
Hello,
On Tue, Jun 06, 2017 at 09:34:05AM -0500, Michael Bringmann wrote:
> >> diff --git a/kernel/workqueue.c b/kernel/workqueue.c
> >> index c74bf39..460de61 100644
> >> --- a/kernel/workqueue.c
> >> +++ b/kernel/workqueue.c
> >> @@ -3366,6 +3366,9 @@ static struct worker_pool
these and other cases, the patch attempts to ensure that a valid,
> usable cpumask is used to set up newly created pools for workqueues.
>
> Signed-off-by: Tejun Heo <t...@kernel.org> & Michael Bringmann
> <m...@linux.vnet.ibm.com>
Heh, you can't add sob's f
Hello, Greg.
On Tue, Mar 14, 2017 at 11:08:29AM +0800, Greg Kroah-Hartman wrote:
> Tejun, want to take this through your tree, or at the least, give me an
> ack for this?
Just acked. I think going through your tree is better for this one.
Thanks!
--
tejun
t_handler+0x1e8/0x1f0
> [c00f1680bdc0] [c00d384c] kthread+0x14c/0x190
> [c00f1680be30] [c000b5a0] ret_from_kernel_thread+0x5c/0xbc
>
> Fixes: f83f3c515654("kernfs: fix locking around kernfs_ops->release()
> callback")
> Signed-off-by: Vaibhav Jain <vaib...@linux.vnet.ibm.com>
Acked-by: Tejun Heo <t...@kernel.org>
Thanks.
--
tejun
Hello, Balbir.
On Tue, Nov 29, 2016 at 11:09:26AM +1100, Balbir Singh wrote:
> On 29/11/16 08:10, Tejun Heo wrote:
> > On Thu, Nov 24, 2016 at 12:05:12AM +1100, Balbir Singh wrote:
> >> On my desktop NODES_SHIFT is 6, many distro kernels have it a 9. I've known
> >>
On Thu, Nov 24, 2016 at 12:05:12AM +1100, Balbir Singh wrote:
> On my desktop NODES_SHIFT is 6, many distro kernels have it a 9. I've known
> of solutions that use fake NUMA for partitioning and need as many nodes as
> possible.
It was a crude kludge that people used before memcg. If people
ch
> adds hotplug support to memory cgroups and reverts the commit that
> limited possible nodes to online nodes.
>
> Cc: Tejun Heo <t...@kernel.org>
> Cc: Andrew Morton <a...@linux-foundation.org>
Can you please cc memcg maintainers? Johannes Weiner
<han...@cmpxch
Hello, Michael.
I simplified the patch a bit and applied the following to for-4.10 and
updated for-next accordingly. The coming next snapshot should be
fine.
Thanks!
-- 8< --
>From 2186d9f940b6a04f263a3bacd48f2a7ba96df4cf Mon Sep 17 00:00:00 2001
From: Tejun Heo <t...@kernel.
Hello, Michael.
On Tue, Oct 18, 2016 at 03:37:42PM +1100, Michael Ellerman wrote:
> That doesn't compile, wq doesn't exist.
>
> I guessed that you meant:
>
> + wq_numa_init();
> + list_for_each_entry(wq, , list)
> + wq_update_unbound_numa(wq, smp_processor_id(), true);
Hello, Michael.
Other NUMA archs are lazy-initializing cpu to node mapping too, so we
need to fix it from workqueue side. This also means that we've been
getting NUMA node wrong for percpu pools on those archs.
Can you please try the following patch and if it resolves the issue,
report the
Hello, Michael.
On Mon, Oct 17, 2016 at 11:24:34PM +1100, Michael Ellerman wrote:
> The bad case (where we hit the BUG_ON I added above) is where we are
> creating a wq for node 1.
>
> In wq_calc_node_cpumask() we do:
>
> cpumask_and(cpumask, attrs->cpumask,
Hello,
On Sat, Oct 15, 2016 at 08:48:01PM +1100, Michael Ellerman wrote:
> > Hmm... it doesn't reproduce it here and can't see how the commit would
> > affect this given that it doesn't really change when the kworker
> > kthreads are being created.
>
> Try turning on CONFIG_DEBUG_PER_CPU_MAPS=y
Hello, Balbir.
On Tue, Oct 11, 2016 at 11:21:09PM +1100, Balbir Singh wrote:
> FYI: I see the samething on my cpu as well, its just that I get lucky
> and cpu_online(cpu) returns false.
Are you seeing this on x86 or is your test setup also a power machine?
Thanks.
--
tejun
Hello, Michael.
On Tue, Oct 11, 2016 at 10:22:13PM +1100, Michael Ellerman wrote:
> The oops happens because we're in enqueue_task_fair() and p->se->cfs_rq
> is NULL.
>
> The cfs_rq is NULL because we did set_task_rq(p, 2048), where 2048 is
> NR_CPUS. That causes us to index past the end of the
On Mon, Oct 10, 2016 at 09:02:53AM -0400, Tejun Heo wrote:
> Hello, Michael.
>
> On Mon, Oct 10, 2016 at 09:22:55PM +1100, Michael Ellerman wrote:
> > This patch seems to be causing one of my Power8 boxes not to boot.
> >
> > Specifically commit 3347fa092821 ("wor
Hello, Michael.
On Mon, Oct 10, 2016 at 09:22:55PM +1100, Michael Ellerman wrote:
> This patch seems to be causing one of my Power8 boxes not to boot.
>
> Specifically commit 3347fa092821 ("workqueue: make workqueue available
> early during boot") in linux-next.
>
> If I revert this on top of
On Mon, Oct 10, 2016 at 10:17:16PM +1100, Balbir Singh wrote:
> rest_init()
> {
> ...
> kernel_thread(kernel_init, NULL, CLONE_FS);
> numa_default_policy();
> pid = kernel_thread(kthreadd, NULL, CLONE_FS | CLONE_FILES);
> rcu_read_lock();
> kthreadd_task =
Hello,
On Thu, Sep 22, 2016 at 02:35:00PM +1000, Nicholas Piggin wrote:
> Well thank you, how about you?
Heh, can't complain. Hope to see you around sometime. It's been
forever.
> Trying a new mail client, sorry. It *seems* to be working now, how's
> this?
Hmm... Still encoded.
> From
Hello, Nick.
How have you been? :)
On Wed, Sep 21, 2016 at 08:57:11PM +1000, Nicholas Piggin wrote:
> On Wed, 21 Sep 2016 18:51:37 +1000
> Nicholas Piggin wrote:
>
> > Some architectures require an additional load to find the address of
> > percpu pointers. In some
e
> increase of local concurrency shouldn't make any difference.
>
> The work item has been flushed in ps3av_remove to ensure that
> there are no pending tasks while disconnecting the driver.
>
> Signed-off-by: Bhaktipriya Shridhar <bhaktipriy...@gmail.com>
Acked-by: Tejun Heo <t...@kernel.org>
Thanks.
--
tejun
On Fri, Jul 15, 2016 at 03:30:41PM +1000, Michael Ellerman wrote:
> It looks like this still hasn't gone to Linus for 4.7?
>
> Could it please, it's a pretty nasty regression on our boxes.
Sorry about that. Just sent out the pull request.
Thanks.
--
tejun
On Tue, Jun 21, 2016 at 09:37:09PM +0200, Peter Zijlstra wrote:
> Hurm.. So I've applied it, just to get this issue sorted, but I'm not
> entirely sure I like it.
>
> I think I prefer ego's version because that makes it harder to get stuff
> to run on !active,online cpus. I think we really want
t; CPU_ONLINE callbacks should be able to put kthreads on the CPU which
> > is coming online. Update select_fallback_rq() so that it follows
> > cpu_online() rather than cpu_active() for kthreads.
> >
> > Signed-off-by: Tejun Heo <t...@kernel.org>
> > Reported-by:
On Thu, Jun 16, 2016 at 02:45:48PM +0200, Peter Zijlstra wrote:
> Subject: workqueue: Fix setting affinity of unbound worker threads
> From: Peter Zijlstra
> Date: Thu Jun 16 14:38:42 CEST 2016
>
> With commit e9d867a67fd03ccc ("sched: Allow per-cpu kernel threads to
> run
rather than cpu_active() for kthreads.
Signed-off-by: Tejun Heo <t...@kernel.org>
Reported-by: Gautham R Shenoy <e...@linux.vnet.ibm.com>
---
kernel/sched/core.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index
Hello,
On Wed, Jun 15, 2016 at 03:14:15PM +0200, Peter Zijlstra wrote:
> On Wed, Jun 15, 2016 at 06:20:33PM +0530, Gautham R Shenoy wrote:
> > We will no longer have the optimization in
> > restore_unbound_workers_cpumask() but I suppose we don't lose much by
> > resetting the affinity every time
Hello,
On Tue, Jun 07, 2016 at 08:44:02PM +0530, Gautham R. Shenoy wrote:
> Currently in the CPU_ONLINE workqueue handler, the
> restore_unbound_workers_cpumask() will never call
> set_cpus_allowed_ptr() for a newly created unbound worker thread.
Hmmm... did you actually verify that this
Hello,
On Tue, May 10, 2016 at 11:34:40AM +0530, Vinod Koul wrote:
> > slave-dma [1], branch topic/dw. But I think Vinod can tell us which
> > tag/branch will be immutable. Vinod?
>
> Please use branch topic/dw. I will not rebase this before sending to Linus.
Okay, pulled topic/dw into
On Mon, May 09, 2016 at 10:13:59AM +0100, Måns Rullgård wrote:
> Andy Shevchenko <andy.shevche...@gmail.com> writes:
>
> > On Mon, May 9, 2016 at 4:09 AM, Tejun Heo <t...@kernel.org> wrote:
> >> On Sun, May 08, 2016 at 04:00:08PM -0400, Tejun Heo wrote:
> >
On Sun, May 08, 2016 at 04:00:08PM -0400, Tejun Heo wrote:
> Hello, Andy.
>
> On Wed, May 04, 2016 at 03:22:51PM +0300, Andy Shevchenko wrote:
> > Tejun, since Vinod applied all necessary patches into his tree, the
> > series now has just a dependency to whatever
Hello, Andy.
On Wed, May 04, 2016 at 03:22:51PM +0300, Andy Shevchenko wrote:
> Tejun, since Vinod applied all necessary patches into his tree, the
> series now has just a dependency to whatever branch / tag he marks for
> it.
> Do we have a chance to see the SATA series be applied in your tree?
On Tue, Apr 26, 2016 at 12:03:04PM +0300, Andy Shevchenko wrote:
> diff --git a/drivers/ata/sata_dwc_460ex.c b/drivers/ata/sata_dwc_460ex.c
> index 7f95389..aee8873 100644
> --- a/drivers/ata/sata_dwc_460ex.c
> +++ b/drivers/ata/sata_dwc_460ex.c
> @@ -1151,7 +1151,13 @@ static struct
Hello,
On Wed, Jul 15, 2015 at 03:43:51PM -0700, Nishanth Aravamudan wrote:
Ok, thank you for clarifying! From a correctness perspective, even if
the numbers don't match NUMA nodes, should we expect the grouping to be
split along NUMA topology?
Yeap, the groups get formed according to the
Hello,
On Fri, Jul 10, 2015 at 09:15:47AM -0700, Nishanth Aravamudan wrote:
On 08.07.2015 [16:16:23 -0700], Nishanth Aravamudan wrote:
On 08.07.2015 [14:00:56 +1000], Michael Ellerman wrote:
On Thu, 2015-02-07 at 23:02:02 UTC, Nishanth Aravamudan wrote:
Much like on x86, now that
Hello,
On Thu, Jul 02, 2015 at 04:02:02PM -0700, Nishanth Aravamudan wrote:
we currently emit at boot:
[0.00] pcpu-alloc: [0] 0 1 2 3 [0] 4 5 6 7
After this commit, we correctly emit:
[0.00] pcpu-alloc: [0] 0 1 2 3 [1] 4 5 6 7
JFYI, the numbers in the brackets aren't
On Tue, May 05, 2015 at 12:14:03PM +0300, Andy Shevchenko wrote:
Since sata_dwc_460ex.c was moved to generic DMA driver we have to ensure that
user can still compile it.
Fixes: 8b3444852a2b (sata_dwc_460ex: move to generic DMA driver)
Signed-off-by: Andy Shevchenko
On Mon, May 04, 2015 at 08:18:30PM +0300, Andy Shevchenko wrote:
On Mon, 2015-04-27 at 23:49 -0400, Tejun Heo wrote:
On Tue, Apr 28, 2015 at 11:12:25AM +1000, Michael Ellerman wrote:
On Mon, 2015-04-27 at 06:59 -0400, Julian Margetson wrote:
Kernel 4.1-rc1 build fails on Sam460ex amcc
On Tue, Apr 28, 2015 at 11:12:25AM +1000, Michael Ellerman wrote:
On Mon, 2015-04-27 at 06:59 -0400, Julian Margetson wrote:
Kernel 4.1-rc1 build fails on Sam460ex amcc 460ex powerpc Canyonlands with
CONFIG_SATA_DWC=y
sata_dwc_460ex.c:(.text+0xa165c): undefined reference to
On Tue, Apr 14, 2015 at 01:58:14PM +0200, Bartlomiej Zolnierkiewicz wrote:
If Tejun Dave are OK with it this patch can go through libata tree.
Otherwise you will need to split it on separate libata/IDE patches.
I really don't care which tree this goes through. Dave, if you wanna
take it,
On Tue, Mar 03, 2015 at 10:41:22PM +0200, Andy Shevchenko wrote:
This patch re-uses hsdev-dev which is allocated on heap. Therefore, the
private structure, which is global variable, is reduced by one field.
In one case ap-dev is used and there it seems to be right decision.
Signed-off-by:
On Thu, Mar 05, 2015 at 10:05:49AM -0800, Nishanth Aravamudan wrote:
While looking at this, I noticed that nr_node_ids is actually a
misnomer, it seems. It's not the number, but the maximum_node_id, as
with sparse NUMA nodes, you might only have two NUMA nodes possible, but
to make certain
On Thu, Mar 05, 2015 at 03:21:35PM -0800, Nishanth Aravamudan wrote:
So, do you agree with the general direction of my change? :)
Yeah, I mean it's an obvious bug fix. I don't know when or how it
should be set on powerpc but if the machine can't do NUMA node
hotplug, its node online and
On Thu, Mar 05, 2015 at 05:08:04PM -0500, Tejun Heo wrote:
It can be argued that we should always stick to the online masks for
allocation and iteration; however, that usually requires more
complexity and the only cases where this mattered have been when the
boot code got it wrong and failed
Hello,
On Thu, Mar 05, 2015 at 01:58:27PM -0800, David Rientjes wrote:
I'm not sure why this is being proposed as a powerpc patch and now a patch
for mem_cgroup_css_alloc(). In other words, why do we have to allocate
for all possible nodes? We should only be allocating for online nodes in
On Thu, Feb 19, 2015 at 03:05:47PM -0500, Martin Hicks wrote:
The driver was ignoring limits requested by libata.force. The output
would look like:
fsl-sata ffe18000.sata: Sata FSL Platform/CSB Driver init
ata1: FORCE: PHY spd limit set to 1.5Gbps
ata1: SATA max UDMA/133 irq 74
ata1:
-by: Kevin Hao haoke...@gmail.com
Acked-by: Tejun Heo t...@kernel.org
Please route as you see fit.
Thanks.
--
tejun
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
Hello,
On Tue, Jul 22, 2014 at 02:43:11PM -0700, Nishanth Aravamudan wrote:
...
There is an issue currently where NUMA information is used on powerpc
(and possibly ia64) before it has been read from the device-tree, which
leads to large slab consumption with CONFIG_SLUB and
On Thu, Jul 17, 2014 at 04:09:23PM -0700, Nishanth Aravamudan wrote:
[Apologies for the large Cc list, but I believe we have the following
interested parties:
x86 (recently posted memoryless node support)
ia64 (existing memoryless node support)
ppc (existing memoryless node support)
Hello,
On Fri, Jul 18, 2014 at 10:42:29AM -0700, Nish Aravamudan wrote:
So, to be clear, this is not *necessarily* about memoryless nodes. It's
about the semantics intended. The workqueue code currently calls
cpu_to_node() in a few places, and passes that node into the core MM as a
hint about
On Fri, Jul 18, 2014 at 02:00:08PM -0400, Tejun Heo wrote:
This isn't a huge issue but it shows that this is the wrong layer to
deal with this issue. Let the allocators express where they are.
^
allocator users
Choosing
Hello,
On Fri, Jul 18, 2014 at 11:12:01AM -0700, Nish Aravamudan wrote:
why aren't these callers using kthread_create_on_cpu()? That API was
It is using that. There just are other data structures too.
already change to use cpu_to_mem() [so one change, rather than of all over
the kernel
Hello,
On Fri, Jul 18, 2014 at 11:47:08AM -0700, Nish Aravamudan wrote:
Why are any callers of the format kthread_create_on_node(...,
cpu_to_node(cpu), ...) not using kthread_create_on_cpu(..., cpu, ...)?
Ah, okay, that's because unbound workers are NUMA node affine, not
CPU.
It seems like
On Tue, Jun 10, 2014 at 04:31:57PM -0700, Nishanth Aravamudan wrote:
I think what this really wants to do is NODE_DATA(cpu_to_mem(cpu)) and I
thought ppc had the cpu-to-local-memory-node mappings correct?
Except cpu_to_mem relies on the mapping being defined, but early in
boot,
Hello,
On Wed, May 21, 2014 at 09:16:27AM -0500, Christoph Lameter wrote:
On Mon, 19 May 2014, Nishanth Aravamudan wrote:
I'm seeing a panic at boot with this change on an LPAR which actually
has no Node 0. Here's what I think is happening:
start_kernel
...
-
-off-by: Tejun Heo t...@kernel.org
Cc: Geoff Levand ge...@infradead.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: cbe-oss-...@lists.ozlabs.org
---
drivers/ps3/ps3-vuart.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/drivers/ps3/ps3-vuart.c b/drivers/ps3/ps3-vuart.c
index fb73008
Hello,
On Thu, Nov 21, 2013 at 9:08 PM, Alistair Popple alist...@popple.id.au wrote:
The new IBM Akebono board has a PPC476GTR SoC with an AHCI compliant
SATA controller. This patch adds a compatible property for the new SoC
to the AHCI platform driver.
Signed-off-by: Alistair Popple
Hello,
On Tue, Oct 08, 2013 at 02:22:16PM +0200, Alexander Gordeev wrote:
If we talk about pSeries quota, then the current pSeries pci_enable_msix()
implementation is racy internally and could fail if the quota went down
*while* pci_enable_msix() is executing. In this case the loop will have
Hello,
On Wed, Oct 09, 2013 at 02:57:16PM +0200, Alexander Gordeev wrote:
On Mon, Oct 07, 2013 at 02:01:11PM -0400, Tejun Heo wrote:
Hmmm... yean, the race condition could be an issue as multiple msi
allocation might fail even if the driver can and explicitly handle
multiple allocation
On Mon, Oct 07, 2013 at 09:48:01PM +0100, Ben Hutchings wrote:
There is one major flaw in min-max approach - the generic MSI layer
will have to take decisions on exact number of MSIs to request, not
device drivers.
[...
No, the min-max functions should be implemented using the same loop
Hello, Alexander.
On Tue, Oct 08, 2013 at 09:48:26AM +0200, Alexander Gordeev wrote:
If there are many which duplicate the above pattern, it'd probably be
worthwhile to provide a helper? It's usually a good idea to reduce
the amount of boilerplate code in drivers.
I wanted to limit
Hello,
On Tue, Oct 08, 2013 at 11:07:16AM +0200, Alexander Gordeev wrote:
Multipe MSIs is just a handful of drivers, really. MSI-X impact still
Yes, so it's pretty nice to try out things there before going full-on.
will be huge. But if we opt a different name for the new pci_enable_msix()
Hey, guys.
On Sun, Oct 06, 2013 at 09:10:30AM +0200, Alexander Gordeev wrote:
On Sun, Oct 06, 2013 at 05:19:46PM +1100, Benjamin Herrenschmidt wrote:
On Sun, 2013-10-06 at 08:02 +0200, Alexander Gordeev wrote:
In fact, in the current design to address the quota race decently the
drivers
Hello,
On Wed, Oct 02, 2013 at 12:48:21PM +0200, Alexander Gordeev wrote:
Make pci_msix_table_size() to return a error code if the device
does not support MSI-X. This update is needed to facilitate a
forthcoming re-design MSI/MSI-X interrupts enabling pattern.
Device drivers will use this
Hello,
On Wed, Oct 02, 2013 at 12:48:23PM +0200, Alexander Gordeev wrote:
+static int foo_driver_enable_msi(struct foo_adapter *adapter, int nvec)
+{
+ rc = pci_get_msi_cap(adapter-pdev);
+ if (rc 0)
+ return rc;
+
+ nvec = min(nvec, rc);
+ if (nvec
Hello, Alexander.
On Wed, Oct 02, 2013 at 12:48:16PM +0200, Alexander Gordeev wrote:
Alexander Gordeev (77):
PCI/MSI: Fix return value when populate_msi_sysfs() failed
PCI/MSI/PPC: Fix wrong RTAS error code reporting
PCI/MSI/s390: Fix single MSI only check
PCI/MSI/s390: Remove
Hello,
On Tue, Oct 01, 2013 at 05:35:48PM +1000, Michael Ellerman wrote:
Roughly third of the drivers just do not care and bail out once
pci_enable_msix() has not succeeded. Not sure how many of these are
mandated by the hardware.
Yeah, I mean, this type of interface is a trap.
On Wed, Oct 02, 2013 at 12:33:38PM +1000, Michael Ellerman wrote:
It is an interface which forces the driver writers to write
complicated fallback code which won't usually be excercised.
It does not force anyone to do anything. That's just bull.
Yeah, sure, we don't have shitty code in
Hello,
On Thu, Sep 26, 2013 at 09:46:46AM +0200, Alexander Gordeev wrote:
Can you please go into a bit of detail on that? Why does it matter?
Because otherwise we will re-introduce a problem described by Michael:
We have a small number of MSIs available, limited by hardware
firmware, if
Hello,
On Thu, Sep 26, 2013 at 10:39 AM, Alexander Gordeev agord...@redhat.com wrote:
I can imagine a scenario where the first device probes in, requests its
Well, we can imagine a lot of thing but usually have to draw the line somewhere.
optimal number, acquires that number and exhausts MSIs
Hello,
On Wed, Sep 25, 2013 at 10:58:05PM +0200, Alexander Gordeev wrote:
Unfortunately, pSeries is a shows-topper here :( It seems we have to
introduce pci_get_msi{,x}_limit() interfaces to honour the quota
thing. I just hope the hardware set for pSeries is limited and we
won't need to use
direct write accesses to using the correct API.
Signed-off-by: Russell King rmk+ker...@arm.linux.org.uk
Acked-by: Tejun Heo t...@kernel.org
The patch is pretty widely spread. I don't mind how it gets routed
but what's the plan?
Thanks.
--
tejun
On Fri, Sep 20, 2013 at 07:16:52AM -0500, Tejun Heo wrote:
On Fri, Sep 20, 2013 at 12:11:38AM +0100, Russell King wrote:
The correct way for a driver to specify the coherent DMA mask is
not to directly access the field in the struct device, but to use
dma_set_coherent_mask(). Only arch
Hello,
On Wed, Sep 18, 2013 at 06:50:45PM +0200, Alexander Gordeev wrote:
Actually, I do not see much contradiction with what I proposed. The
key words here determine the number of MSIs the controller wants.
In general case it is not what pci_msix_table_size() returns (or at
least we should
On Fri, Sep 20, 2013 at 10:24:59AM +0200, Alexander Gordeev wrote:
* Make pci_enable_msix() return 0/-errno
My choice would be this one.
Thanks.
--
tejun
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
Hey,
On Fri, Sep 20, 2013 at 03:00:18PM +0100, Russell King - ARM Linux wrote:
Another would be if subsystem maintainers are happy that I carry them,
I can add the acks, and then later on towards the end of the cycle,
provide a branch subsystem maintainers could pull.
Or... if you can think
Hello,
On Wed, Sep 18, 2013 at 11:48:00AM +0200, Alexander Gordeev wrote:
On Wed, Sep 18, 2013 at 12:30:23AM +1000, Michael Ellerman wrote:
How about no?
We have a small number of MSIs available, limited by hardware
firmware, if we don't impose a quota then the first device that probes
On Mon, Aug 19, 2013 at 07:20:30PM -0600, Anthony Foiani wrote:
This patch is based off linux-next tag next-20130819
(which is commit 66a01bae29d11916c09f9f5a937cafe7d402e4a5 )
Signed-off-by: Anthony Foiani anthony.foi...@gmail.com
Applied to libata/for-3.11-fixes w/ stable cc'd.
Thanks.
On Wed, Jun 26, 2013 at 10:51:11AM -0400, Steven Rostedt wrote:
It would also increase the latency of CPU-hotunplug operations.
Is that a big deal?
I thought that was the whole deal with this patchset - making cpu
hotunplugs lighter and faster mostly for powersaving. That said, just
Hello,
On Wed, Jun 26, 2013 at 11:33:43AM -0400, Steven Rostedt wrote:
I thought the whole deal with this patchset was to remove stop_machine
from CPU hotplug. Why halt all CPUs just to remove one? stomp_machine()
is extremely intrusive for the entire system, where as one CPU making
sure all
Hey,
On Wed, Jun 26, 2013 at 11:58:48PM +0530, Srivatsa S. Bhat wrote:
Yes, we were discussing hot-unplug latency for use-cases such as
suspend/resume. We didn't want to make those operations slower in the
process of removing stop_machine() from hotplug.
Can you please explain why tho? How
On Sun, Jun 23, 2013 at 07:12:59PM +0530, Srivatsa S. Bhat wrote:
Once stop_machine() is gone from the CPU offline path, we won't be able
to depend on disabling preemption to prevent CPUs from going offline
from under us.
Use the get/put_online_cpus_atomic() APIs to prevent CPUs from going
1 - 100 of 150 matches
Mail list logo