On Fri, 2017-04-07 at 22:55 +1000, Nicholas Piggin wrote:
> POWER9 hypervisors will not necessarily run guest threads together on
> the same core at the same time, so msgsndp should not be used.
Maybe we shouldn't advertise doorbells at all ?
> Signed-off-by: Nicholas Piggin
> ---
> arch/powerp
Hi! Find below my third regression report for Linux 4.11. It lists 15
regressions I'm currently aware of. 5 regressions mentioned in last
weeks report got fixed.
As always: Are you aware of any other regressions? Then please let me
know (simply CC regressi...@leemhuis.info). And please tell me if
On Sun, Apr 09, 2017 at 07:15:49PM +0200, Thorsten Leemhuis wrote:
> Desc: i915 gpu hangs under load
> Repo: 2017-03-22
> https://www.mail-archive.com/intel-gfx@lists.freedesktop.org/msg116227.html
> https://bugs.freedesktop.org/show_bug.cgi?id=100181
> https://www.mail-archive.com/linux-kernel@
Current behaviour assumes that memory in RAM is contiguous and
iterates from the start of RAM to (start + size of memory). When the
memory isn't physically contiguous, this approach doesn't work.
If memory exists at 0-5 GB and 6-10 GB then the current approach will
check if entries exist in the ha
mempool_alloc() cannot fail when passed GFP_NOIO or any
other gfp setting that is permitted to sleep.
So remove this pointless code.
Signed-off-by: NeilBrown
---
drivers/scsi/ibmvscsi/ibmvfc.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi
On Sat, 1 Apr 2017 20:11:48 +0530
"Aneesh Kumar K.V" wrote:
> For a tlbiel with pid, we need to issue tlbiel with set number
> encoded. We don't need to do ptesync for each of those. Instead we
> need one for the entire tlbiel pid operation.
>
> Signed-off-by: Benjamin Herrenschmidt
> Signed-o
On Sat, 1 Apr 2017 20:11:47 +0530
"Aneesh Kumar K.V" wrote:
> For fullmm tlb flush, we do a flush with RIC_FLUSH_ALL which will
> invalidate all related caches (radix__tlb_flush()). Hence the pwc
> flush is not needed.
Thanks Aneesh. I see a 3x improvement in exec performance with these
2 patch
On Sun, 09 Apr 2017 18:03:35 +1000
Benjamin Herrenschmidt wrote:
> On Fri, 2017-04-07 at 22:55 +1000, Nicholas Piggin wrote:
> > POWER9 hypervisors will not necessarily run guest threads together on
> > the same core at the same time, so msgsndp should not be used.
>
> Maybe we shouldn't adver
On Friday 07 April 2017 06:06 PM, Michael Ellerman wrote:
Sachin Sant writes:
I have run into few instances where the lost_exception_test from
powerpc kselftest fails with SIGABRT. Following o/p is against
4.11.0-rc5. The failure is intermittent.
What hardware are you on?
How long does it
Reviewed-by: Andrew Donnellan
On 08/04/17 00:11, Christophe Lombard wrote:
This bit is used to cause a flash image load for programmable
CAIA-compliant implementation. If this bit is set to ‘0’, a power
cycle of the adapter is required to load a programmable CAIA-com-
pliant implementation from
On 04/10/2017 07:34 AM, Rashmica Gupta wrote:
> Current behaviour assumes that memory in RAM is contiguous and
> iterates from the start of RAM to (start + size of memory). When the
> memory isn't physically contiguous, this approach doesn't work.
In dis contiguous platform, we just finish up walk
On Mon, 2017-04-10 at 13:22 +1000, Nicholas Piggin wrote:
> On Sun, 09 Apr 2017 18:03:35 +1000
> > Benjamin Herrenschmidt wrote:
>
> > On Fri, 2017-04-07 at 22:55 +1000, Nicholas Piggin wrote:
> > > POWER9 hypervisors will not necessarily run guest threads together on
> > > the same core at the s
On Mon, 10 Apr 2017 14:07:38 +1000
Benjamin Herrenschmidt wrote:
> On Mon, 2017-04-10 at 13:22 +1000, Nicholas Piggin wrote:
> > On Sun, 09 Apr 2017 18:03:35 +1000
> > > Benjamin Herrenschmidt wrote:
> >
> > > On Fri, 2017-04-07 at 22:55 +1000, Nicholas Piggin wrote:
> > > > POWER9 hype
On 10/04/17 14:02, Anshuman Khandual wrote:
On 04/10/2017 07:34 AM, Rashmica Gupta wrote:
Current behaviour assumes that memory in RAM is contiguous and
iterates from the start of RAM to (start + size of memory). When the
memory isn't physically contiguous, this approach doesn't work.
In dis
The patch to introduce address translation services for Nvlink2 uses
MMU notifiers. However usage of MMU notifiers requires a Kconfig
option which is not selected by default on powerpc so add it to the
powernv Kconfig.
Signed-off-by: Alistair Popple
---
arch/powerpc/platforms/powernv/Kconfig | 1
On 08/04/17 00:11, Christophe Lombard wrote:
The two previously fields pid and tid, located in the structure
cxl_irq_info, are only used in the guest environment. To avoid confusion,
it's not necessary to fill the fields in the bare-metal environment.
Pid_tid is now renamed to 'reserved' to avoid
On Thu, Apr 6, 2017 at 4:37 PM, Oliver O'Halloran wrote:
> On Thu, Mar 30, 2017 at 10:28 AM, Matt Brown
> wrote:
>> The HDAT data area is consumed by skiboot and turned into a device-tree. In
>> some cases we would like to look directly at the HDAT. This is not possible
>> through /dev/mem as it
On 08/04/17 00:11, Christophe Lombard wrote:
The mm_struct corresponding to the current task is acquired each time
an interrupt is raised. So to simplify the code, we only get the
mm_struct when attaching an AFU context to the process.
The mm_count reference is increased to ensure that the mm_str
On Thu, 6 Apr 2017 09:09:41 -0500
Rob Herring wrote:
> On Wed, Apr 5, 2017 at 7:38 PM, Nicholas Piggin wrote:
> > Given that it's quite a small addition to of/fdt code, hopefully
> > that gives you a reasonable justification to accept it.
> >
> > If you prefer not to, that's okay, but I think we
Dumping the PE State Tables (PEST) can be highly verbose if a number of PEs
are affected, especially in the case where the whole PHB is frozen and 255
lines get printed. Check for duplicates when dumping the PEST to reduce
useless output.
For example:
PE[f8] A/B: 9726 8080d00
On 08/04/17 00:11, Christophe Lombard wrote:
Rename a few functions, changing the '_psl' suffix to '_psl8', to make
clear that the implementation is psl8 specific.
Those functions will have an equivalent implementation for the psl9 in
a later patch.
Signed-off-by: Christophe Lombard
Reviewed-
If the interrupt didn't have a selected target yet, we could
call cpu_online() and do other cpumask tests with cpu #-1 which
would result in random outcomes.
Signed-off-by: Benjamin Herrenschmidt
---
arch/powerpc/sysdev/xive/common.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff
When targetting interrupts we do various manipulations of cpu numbers
and CPU masks. This adds some sanity checking to ensure we don't
break assumptions and manpulate cpu numbers that are out of bounds
of the various cpumasks.
Signed-off-by: Benjamin Herrenschmidt
---
arch/powerpc/sysdev/xive/co
xive_pick_irq_target() first tries to construct a mask that is
the intersection of the requested affinity, online CPUs, and
the group of CPUs that are on the same chip as the interrupt
source.
If that resulting mask is empty, we were incorrectly returning
nr_cpu_ids as a target.
Signed-off-by: Be
If the interrupt didn't have a selected target yet, we could
call cpu_online() and do other cpumask tests with cpu #-1 which
would result in random outcomes.
Signed-off-by: Benjamin Herrenschmidt
---
arch/powerpc/sysdev/xive/common.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff
xive_pick_irq_target() first tries to construct a mask that is
the intersection of the requested affinity, online CPUs, and
the group of CPUs that are on the same chip as the interrupt
source.
If that resulting mask is empty, we were incorrectly returning
nr_cpu_ids as a target.
Signed-off-by: Be
When targetting interrupts we do various manipulations of cpu numbers
and CPU masks. This adds some sanity checking to ensure we don't
break assumptions and manpulate cpu numbers that are out of bounds
of the various cpumasks.
Signed-off-by: Benjamin Herrenschmidt
---
arch/powerpc/sysdev/xive/co
Daniel Axtens writes:
> Hi Matt,
>
> Thanks for answering my questions and doing those fixes.
>
>
>> Bugs fixed:
>> - A small bug in pq.h regarding a missing and mismatched
>>ifdef statement
>> - Fixed test/Makefile to correctly build test on ppc
>>
>
> I think this commit shoul
28 matches
Mail list logo
