[Added KVM lists and a couple relevant people]
On Fri, 2015-07-31 at 14:25 +0530, Hemant Kumar wrote:
> On 07/30/2015 03:52 AM, Scott Wood wrote:
> > On Wed, 2015-07-29 at 16:07 +0530, Hemant Kumar wrote:
> > > Hi Scott,
> > >
> > > On 07/17/2015 01:40 AM, Scott Wood wrote:
> > > > On Thu, 2015-0
On 31.7.2015 23:25, David Rientjes wrote:
> On Thu, 30 Jul 2015, Vlastimil Babka wrote:
>
>> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
>> index aa58a32..56355f2 100644
>> --- a/mm/huge_memory.c
>> +++ b/mm/huge_memory.c
>> @@ -2469,7 +2469,7 @@ khugepaged_alloc_page(struct page **hpage, gfp
On Thu, 30 Jul 2015, Vlastimil Babka wrote:
> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> index aa58a32..56355f2 100644
> --- a/mm/huge_memory.c
> +++ b/mm/huge_memory.c
> @@ -2469,7 +2469,7 @@ khugepaged_alloc_page(struct page **hpage, gfp_t gfp,
> struct mm_struct *mm,
>*/
>
3.14-stable review patch. If anyone has any objections, please let me know.
--
From: Nicholas Mc Guire
commit ed9244e6c534612d2b5ae47feab2f55a0d4b4ced upstream.
Fix possible unintended sign extension in unsigned MMIO loads by casting
to uint16_t in the case of mmio_needed != 2
3.10-stable review patch. If anyone has any objections, please let me know.
--
From: Nicholas Mc Guire
commit ed9244e6c534612d2b5ae47feab2f55a0d4b4ced upstream.
Fix possible unintended sign extension in unsigned MMIO loads by casting
to uint16_t in the case of mmio_needed != 2
On Fri, Jul 31, 2015 at 04:54:01PM +0100, Alex Bennée wrote:
> From: Alex Bennée
>
> This test has been written mainly to stress multi-threaded TCG behaviour
> but will demonstrate failure by default on real hardware. The test takes
> the following parameters:
>
> - "lock" use GCC's locking se
On Fri, Jul 31, 2015 at 04:53:59PM +0100, Alex Bennée wrote:
> From: Alex Bennée
>
> The previous $(getconf _NPROCESSORS_CONF) isn't correct as the default
> maximum VCPU configuration is 4 on arm64 machines which typically have
> more actual cores.
The kernel should probably bump that up to 8,
On Fri, Jul 31, 2015 at 04:53:58PM +0100, Alex Bennée wrote:
> ---
> arm/unittests.cfg | 24
> 1 file changed, 24 insertions(+)
>
> diff --git a/arm/unittests.cfg b/arm/unittests.cfg
> index ee655b2..19d72ad 100644
> --- a/arm/unittests.cfg
> +++ b/arm/unittests.cfg
> @@
On Fri, Jul 31, 2015 at 04:53:57PM +0100, Alex Bennée wrote:
> This adds a fairly brain dead torture test for TLB flushes intended for
> stressing the MTTCG QEMU build. It takes the usual -smp option for
> multiple CPUs.
>
> By default it CPU0 will do a TLBIALL flush after each cycle. You can
> pa
On Fri, Jul 31, 2015 at 04:53:56PM +0100, Alex Bennée wrote:
> This introduces a new flush_tlb_page function which does exactly what
> you expect. It's going to be useful for the future TLB torture test.
>
> Signed-off-by: Alex Bennée
> ---
> lib/arm/asm/mmu.h | 11 +++
> lib/arm64/asm
On Fri, Jul 31, 2015 at 04:53:55PM +0100, Alex Bennée wrote:
> ---
> lib/printf.c | 13 +
> 1 file changed, 13 insertions(+)
missing sign-off
>
> diff --git a/lib/printf.c b/lib/printf.c
> index 89308fb..5d83605 100644
> --- a/lib/printf.c
> +++ b/lib/printf.c
> @@ -180,6 +180,19 @@
Hi Paolo,
Thank you for the answer. I suspect that overcommit thing. In my case
it's a combination of unused memory and KSM.
So hugepages are not so useful as I saw in the presentations. I think
it works just for some usecases.
Thanks
On Fri, Jul 31, 2015 at 12:54 AM, Paolo Bonzini wrote:
>
>
This makes the script a little cleaner by only checking for KVM support
in one place. If KVM isn't available we can fall back to TCG emulation
and echo the fact to the screen rather than let QEMU complain.
Signed-off-by: Alex Bennée
Reviewed-by: Andrew Jones
---
v2
- rm redundant M= statement
From: Alex Bennée
The previous $(getconf _NPROCESSORS_CONF) isn't correct as the default
maximum VCPU configuration is 4 on arm64 machines which typically have
more actual cores.
This introduces a simple utility program to query the KVM capabilities
and use the correct maximum number of vcpus.
Hi,
This is the current state of my MTTCG tests based on the KVM's unit
testing framework. The earlier patches in the series have already been
reviewed and will (with the exception of the emacs patch) be making
their way upstream.
There are a couple of addition to library functions:
- printf %u
This is useful information for the run scripts to know, especially if
they want to drop to using TCG.
Signed-off-by: Alex Bennée
Reviewed-by: Andrew Jones
---
v3
- add r-b tag
---
configure | 2 ++
1 file changed, 2 insertions(+)
diff --git a/configure b/configure
index b2ad32a..078b70c 100
This introduces a new flush_tlb_page function which does exactly what
you expect. It's going to be useful for the future TLB torture test.
Signed-off-by: Alex Bennée
---
lib/arm/asm/mmu.h | 11 +++
lib/arm64/asm/mmu.h | 8
2 files changed, 19 insertions(+)
diff --git a/lib/a
---
arm/unittests.cfg | 24
1 file changed, 24 insertions(+)
diff --git a/arm/unittests.cfg b/arm/unittests.cfg
index ee655b2..19d72ad 100644
--- a/arm/unittests.cfg
+++ b/arm/unittests.cfg
@@ -35,3 +35,27 @@ file = selftest.flat
smp = $(getconf _NPROCESSORS_CONF)
extra
This adds a fairly brain dead torture test for TLB flushes intended for
stressing the MTTCG QEMU build. It takes the usual -smp option for
multiple CPUs.
By default it CPU0 will do a TLBIALL flush after each cycle. You can
pass options via -append to control additional aspects of the test:
- "p
Signed-off-by: Alex Bennée
Reviewed-by: Andrew Jones
---
v2
- mention consistency
v3
- add r-b tag
---
README | 26 ++
1 file changed, 26 insertions(+)
diff --git a/README b/README
index e9869d1..9389a26 100644
--- a/README
+++ b/README
@@ -25,3 +25,29 @@ Directory
---
lib/printf.c | 13 +
1 file changed, 13 insertions(+)
diff --git a/lib/printf.c b/lib/printf.c
index 89308fb..5d83605 100644
--- a/lib/printf.c
+++ b/lib/printf.c
@@ -180,6 +180,19 @@ int vsnprintf(char *buf, int size, const char *fmt,
va_list va)
break;
Signed-off-by: Alex Bennée
---
arm/run | 1 +
1 file changed, 1 insertion(+)
diff --git a/arm/run b/arm/run
index 662a856..6b42a2e 100755
--- a/arm/run
+++ b/arm/run
@@ -1,4 +1,5 @@
#!/bin/bash
+# -*- sh-basic-offset:8 indent-tabs-mode: t -*-
if [ ! -f config.mak ]; then
echo run ./c
From: Alex Bennée
This test has been written mainly to stress multi-threaded TCG behaviour
but will demonstrate failure by default on real hardware. The test takes
the following parameters:
- "lock" use GCC's locking semantics
- "excl" use load/store exclusive semantics
- "acqrel" use acqu
On 31/07/2015 17:06, Alex Bennée wrote:
>
> I haven't seen this upstream yet. Has this been pushed yet?
Nope...
Paolo
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo
Paolo Bonzini writes:
> Applied as follows:
>
>> +Maintainers
>> +---
>> +M: Paolo Bonzini
>> +M: Marcelo Tosatti
>> +L: kvm@vger.kernel.org
>> +T: git://git.kernel.org/pub/scm/virt/kvm/kvm-unit-tests.git
>> +
>> +Architecture Specific Code:
>> +---
>> +
>> +ARM
Hi Andre,
On 07/10/2015 04:21 PM, Andre Przywara wrote:
> When userland wants to inject a MSI into the guest, we have to use
> our data structures to find the LPI number and the VCPU to receive
> the interrupt.
> Use the wrapper functions to iterate the linked lists and find the
> proper Interrupt
Hi Andre,
On 07/11/2015 01:17 AM, Andre Przywara wrote:
> On 09/07/15 09:22, Eric Auger wrote:
>> If the ITS modality is not available, let's simply support MSI
>> injection by transforming the MSI.data into an SPI ID.
>>
>> This becomes possible to use KVM_SIGNAL_MSI ioctl for arm too.
>>
>> Signe
Test the undocumented ICEBP (aka INT1 or INT01) opcode, and test that
dr6.BS is not modified by neither watchpoints nor breakpoints.
Signed-off-by: Paolo Bonzini
---
x86/debug.c | 51 ---
1 file changed, 44 insertions(+), 7 deletions(-)
diff --git
On 2015-07-30 23:19, Steve Rutherford wrote:
> On Thu, Jul 30, 2015 at 11:38:20AM +0200, Paolo Bonzini wrote:
>>
>>
>> On 30/07/2015 10:37, Steve Rutherford wrote:
>>> This looks a bit non-sensical, but is overprepared for the introduction
>>> IOAPIC hotplug, which two patches down the line. Changi
On 31/07/2015 01:26, Zhang, Yang Z wrote:
>>> Do not compute TMR in advance. Instead, set the TMR just before
>>> the interrupt is accepted into the IRR. This limits the coupling
>>> between IOAPIC and LAPIC.
>
> Uh.., it back to original way which is wrong. You cannot modify the
> apic page(h
On 31/07/2015 04:49, Steve Rutherford wrote:
> Oh... Yeah. That's a damn good point, given that the interrupt can be
> injected from another thread while one is in that guest vcpu.
>
> Easiest time to update the TMR should be on guest entry through
> vcpu_scan_ioapic, as before.
>
> The best wa
On 31/07/2015 01:08, Pavel Shirshov wrote:
> Hello,
>
> Today I was very surprised to see that KVM with enabled HugePages uses
> much more memory versus KVM without HugePages enabled.
>
> I have a server with 386Gb memory.
> I have a VM image which allocate up to 3.2Gb memory (libvirt tag).
>
Hi David,
As noticed we did the same on ARM (4.2) for the same reason.
Reviewed-by: Eric Auger
Best Regards
Eric
On 07/20/2015 08:12 AM, David Gibson wrote:
> ec53500f "kvm: Add VFIO device" added a special KVM pseudo-device which is
> used to handle any necessary interactions between KVM an
33 matches
Mail list logo
