On 4/3/08, Anthony Liguori <[EMAIL PROTECTED]> wrote:
> Jun Koi wrote:
>
> > Hi,
> >
> > Could someone please post some concret examples on how to use extboot?
> > I looked around, but saw nothing.
> >
> >
>
> Just append boot=on to your -drive parameter.
>
OK, but the question is that when I nee
Hi Rusty,
I hit a bug in virtio_ring.c:218 when I was stressing virtio_net using kvm with
-smp 4.
static void vring_disable_cb(struct virtqueue *_vq)
{
struct vring_virtqueue *vq = to_vvq(_vq);
START_USE(vq);
-->BUG_ON(vq->vring.avail->flags & VRING_AVAIL_F_NO_INTERRUPT)
Le vendredi 04 avril 2008 à 18:29 +0900, Jun Koi a écrit :
> On 4/3/08, Anthony Liguori <[EMAIL PROTECTED]> wrote:
> > Jun Koi wrote:
> >
> > > Hi,
> > >
> > > Could someone please post some concret examples on how to use extboot?
> > > I looked around, but saw nothing.
> > >
> > >
> >
> > Just a
On 4/4/08, Laurent Vivier <[EMAIL PROTECTED]> wrote:
>
> Le vendredi 04 avril 2008 à 18:29 +0900, Jun Koi a écrit :
>
> > On 4/3/08, Anthony Liguori <[EMAIL PROTECTED]> wrote:
> > > Jun Koi wrote:
> > >
> > > > Hi,
> > > >
> > > > Could someone please post some concret examples on how to use
On Thu, Apr 03, 2008 at 12:20:41PM -0700, Christoph Lameter wrote:
> On Thu, 3 Apr 2008, Andrea Arcangeli wrote:
>
> > My attempt to fix this once and for all is to walk all vmas of the
> > "mm" inside mmu_notifier_register and take all anon_vma locks and
> > i_mmap_locks in virtual address order
From: Heiko Carstens <[EMAIL PROTECTED]>
Just a few codingstyle issues that have already been
commented on but that somehow got forgotten.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Carsten Otte <[EMAIL PROTECTED]>
---
arch/s390/kvm/gaccess.h | 18 ++
ar
From: Heiko Carstens <[EMAIL PROTECTED]>
Temporarily rename this function to avoid merge conflicts and/or
dependencies. This function will be removed as soon as git-s390
and kvm.git are finally upstream.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Carsten Otte <[EMAIL PROTECT
From: Heiko Carstens <[EMAIL PROTECTED]>
kvm_arch_vcpu_ioctl_run currently incorrectly always returns 0.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Carsten Otte <[EMAIL PROTECTED]>
---
arch/s390/kvm/kvm-s390.c |2 +-
1 files changed, 1 insertion(+), 1 deletion(-)
Index
From: Heiko Carstens <[EMAIL PROTECTED]>
There is no need to use interlocked updates when the rcp
lock is held. Therefore the simple bitops variants can be
used. This should improve performance.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Carsten Otte <[EMAIL PROTECTED]>
---
Hi Avi,
looks like Heiko played around with kvm a bit :-):
01/04: cosmetics, including part of Arnd's review feedback
02/04: bugfix, if KVM_VCPU_IOCTL_RUN got interrupted by signal
03/04: performance improvement with our pgste memory handling
04/04: resolve merge conflict in linux-next kvm.git ver
Pamper yourself to the maximum
Get ready to live the luxurious life with your new collection of items
http://www.kerinaes.com/
-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
On Fri, 2008-04-04 at 15:12 +0200, Carsten Otte wrote:
> Index: kvm/include/asm-s390/pgtable.h
> ===
> --- kvm.orig/include/asm-s390/pgtable.h
> +++ kvm/include/asm-s390/pgtable.h
> @@ -553,12 +553,12 @@ static inline void ptep_rcp_cop
Jun Koi wrote:
> On 4/4/08, Laurent Vivier <[EMAIL PROTECTED]> wrote:
>
>> Le vendredi 04 avril 2008 à 18:29 +0900, Jun Koi a écrit :
>>
>>
>>> On 4/3/08, Anthony Liguori <[EMAIL PROTECTED]> wrote:
>>>
>> > > Jun Koi wrote:
>> > >
>> > > > Hi,
>> > > >
>> > > > Could someone pl
From: Heiko Carstens <[EMAIL PROTECTED]>
There is no need to use interlocked updates when the rcp
lock is held. Therefore the simple bitops variants can be
used. This should improve performance.
Signed-off-by: Heiko Carstens <[EMAIL PROTECTED]>
Signed-off-by: Carsten Otte <[EMAIL PROTECTED]>
---
[EMAIL PROTECTED] wrote:
> Major formatting accident ?
Ops! How did that happen? Thanks, will resend that patch.
-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just abo
on Wed Apr 02 2008, Avi Kivity wrote:
> David Abrahams wrote:
>> With the title combination, the guest takes nearly 100% of my real CPU
>> time and still only sees one CPU. Is this a known problem, and does it
>> have a known solution?
>>
>>
>
> Can you send the output of 'kvm_stat -1'?
$
* Marcelo Tosatti <[EMAIL PROTECTED]> [2008-04-03 18:22]:
>
> kvm_arch_load_regs() might load APIC base and APIC TPR from the
> uninitialized userspace APIC emulation.
>
> This happened when running X11 on SMP guest in my case, disabling the
> APIC for vcpu3:
>
> vcpu not ready for apic_round_ro
Zdenek reported a bug where a looping "dmsetup status" eventually hangs
on SMP guests.
The problem is that kvm_mmu_get_page() prepopulates the shadow MMU
before write protecting the guest page tables. By doing so, it leaves a
window open where the guest can mark a pte as present while the host ha
2008/4/4, Marcelo Tosatti <[EMAIL PROTECTED]>:
>
> Zdenek reported a bug where a looping "dmsetup status" eventually hangs
> on SMP guests.
>
> The problem is that kvm_mmu_get_page() prepopulates the shadow MMU
> before write protecting the guest page tables. By doing so, it leaves a
> window
Account for the case where memslot size is largepage aligned, but does not
start on largepage aligned offset.
Otherwise it allocates one lpage_info struct less than what the memslot
spans, possibly resulting in memory corruption.
Signed-off-by: Marcelo Tosatti <[EMAIL PROTECTED]>
diff --git
This should guarantee that nobody can register when any of the mmu
notifiers is running avoiding all the races including guaranteeing
range_start not to be missed. I'll adapt the other patches to provide
the sleeping-feature on top of this (only needed by XPMEM) soon. KVM
seems to run fine on top o
Marcelo Tosatti wrote:
> Zdenek reported a bug where a looping "dmsetup status" eventually hangs
> on SMP guests.
>
> The problem is that kvm_mmu_get_page() prepopulates the shadow MMU
> before write protecting the guest page tables. By doing so, it leaves a
> window open where the guest can mark a
Carsten Otte wrote:
> From: Heiko Carstens <[EMAIL PROTECTED]>
>
> There is no need to use interlocked updates when the rcp
> lock is held. Therefore the simple bitops variants can be
> used. This should improve performance.
>
> skey = page_get_storage_key(page_to_phys(page));
> if (skey & _PAGE_CH
Carsten Otte wrote:
> Hi Avi,
>
> looks like Heiko played around with kvm a bit :-):
> 01/04: cosmetics, including part of Arnd's review feedback
> 02/04: bugfix, if KVM_VCPU_IOCTL_RUN got interrupted by signal
> 03/04: performance improvement with our pgste memory handling
> 04/04: resolve merge c
Carsten Otte wrote:
> From: Heiko Carstens <[EMAIL PROTECTED]>
>
> There is no need to use interlocked updates when the rcp
> lock is held. Therefore the simple bitops variants can be
> used. This should improve performance.
>
>
Applied this fixed patch as well, thanks.
--
Any sufficiently dif
Anthony Liguori wrote:
> The vmport code is very broken for SMP guests. It uses a global CPUState
> that's initialized multiple times? At any rate, since it needs to know CPU
> registers for the current CPU in a PIO handler, it needs to use
> cpu_single_env.
>
> This patch makes vmmouse when usi
Marcelo Tosatti wrote:
> kvm_arch_load_regs() might load APIC base and APIC TPR from the
> uninitialized userspace APIC emulation.
>
> This happened when running X11 on SMP guest in my case, disabling the
> APIC for vcpu3:
>
> vcpu not ready for apic_round_robin
> vcpu not ready for apic_round_robi
Marcelo Tosatti wrote:
> Account for the case where memslot size is largepage aligned, but does not
> start on largepage aligned offset.
>
> Otherwise it allocates one lpage_info struct less than what the memslot
> spans, possibly resulting in memory corruption.
>
Applied, thanks.
--
Any suf
KVM kernel modules are now located in arch/x86/kvm so make sure to rename those
too.
Signed-off-by: Anthony Liguori <[EMAIL PROTECTED]>
diff --git a/kernel/Makefile b/kernel/Makefile
index fb053eb..c30c8b2 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -61,7 +61,8 @@ sync:
install:
I am always the guy doing the cleanup after Andrea it seems. Sigh.
Here is the mm_lock/mm_unlock logic separated out for easier review.
Adds some comments. Still objectionable is the multiple ways of
invalidating pages in #v11. Callout now has similar locking to emm.
From: Christoph Lameter <[EMA
The conversion to a rwsem allows callbacks during rmap traversal
for files in a non atomic context. A rw style lock also allows concurrent
walking of the reverse map. This is fairly straightforward if one removes
pieces of the resched checking.
[Restarting unmapping is an issue to be discussed].
Provide a way to lock an mm_struct against reclaim (try_to_unmap
etc). This is necessary for the invalidate notifier approaches so
that they can reliably add and remove a notifier.
Signed-off-by: Andrea Arcangeli <[EMAIL PROTECTED]>
Signed-off-by: Christoph Lameter <[EMAIL PROTECTED]>
---
includ
This patch implements a simple callback for device drivers that establish
their own references to pages (KVM, GRU, XPmem, RDMA/Infiniband, DMA engines
etc). These references are unknown to the VM (therefore external).
With these callbacks it is possible for the device driver to release external
re
Convert the anon_vma spinlock to a rw semaphore. This allows concurrent
traversal of reverse maps for try_to_unmap and page_mkclean. It also
allows the calling of sleeping functions from reverse map traversal.
An additional complication is that rcu is used in some context to guarantee
the presence
A simple test program (well actually a pair). They are fairly easy to use.
NOTE: the xpmem.h is copied from the kernel/drivers/misc/xp/xpmem.h
file.
Type make. Then from one session, type ./A1. Grab the first
line of output which should begin with ./A2 and paste the whole line
into a second se
V2->V3:
- Fix rcu issues
- Fix emm_referenced handling
- Use Andrea's mm_lock/unlock to prevent registration races.
- Keep simple API since there does not seem to be a need to add additional
callbacks (mm_lock does not require callbacks like emm_start/stop that
I envisioned).
- Reduce CC list (
Move the tlb flushing into free_pgtables. The conversion of the locks
taken for reverse map scanning would require taking sleeping locks
in free_pgtables(). Moving the tlb flushing into free_pgtables allows
sleeping in parts of free_pgtables().
This means that we do a tlb_finish_mmu() before freei
We no longer abort unmapping in unmap vmas because we can reschedule while
unmapping since we are holding a semaphore. This would allow moving more
of the tlb flusing into unmap_vmas reducing code in various places.
Signed-off-by: Christoph Lameter <[EMAIL PROTECTED]>
---
include/linux/mm.h |
XPMEM would have used sys_madvise() except that madvise_dontneed()
returns an -EINVAL if VM_PFNMAP is set, which is always true for the pages
XPMEM imports from other partitions and is also true for uncached pages
allocated locally via the mspec allocator. XPMEM needs zap_page_range()
functionalit
This patch adds a lock ordering rule to avoid a potential deadlock when
multiple mmap_sems need to be locked.
Signed-off-by: Dean Nelson <[EMAIL PROTECTED]>
---
mm/filemap.c |3 +++
1 file changed, 3 insertions(+)
Index: linux-2.6/mm/filemap.c
===
Christoph Lameter wrote:
> Provide a way to lock an mm_struct against reclaim (try_to_unmap
> etc). This is necessary for the invalidate notifier approaches so
> that they can reliably add and remove a notifier.
>
> Signed-off-by: Andrea Arcangeli <[EMAIL PROTECTED]>
> Signed-off-by: Christoph Lame
Hi Folks,
A fellow student and I wish to do run some experiments with KVM.
Specifically, we would like to examine
KVM's guest paging/swapping mechanism, make some changes and run some tests.
After a brief search through
the docs and code, we would greatly appreciate help with the following:
1) Gu
On Fri, Apr 04, 2008 at 03:06:18PM -0700, Christoph Lameter wrote:
> Adds some comments. Still objectionable is the multiple ways of
> invalidating pages in #v11. Callout now has similar locking to emm.
range_begin exists because range_end is called after the page has
already been freed. invalidat
On Fri, Apr 04, 2008 at 04:12:42PM -0700, Jeremy Fitzhardinge wrote:
> I think you can break this if() down a bit:
>
> if (!(vma->vm_file && vma->vm_file->f_mapping))
> continue;
It makes no difference at runtime, coding style preferences are qui
On Fri, Apr 04, 2008 at 03:30:50PM -0700, Christoph Lameter wrote:
> + mm_lock(mm);
> + e->next = mm->emm_notifier;
> + /*
> + * The update to emm_notifier (e->next) must be visible
> + * before the pointer becomes visible.
> + * rcu_assign_pointer() does exactly what we
Dor Laor wrote:
>
>> +static uint32_t vring_desc_len(VirtQueue *vq, unsigned int i)
>> +{
>
> Below there were place you did use offsetof(vq->vring.desc[i], len) so
> we better be consistent + its nicer
>
>> +return ldl_phys(vq->vring.desc + i * sizeof(VRingDesc) +
>> +offse
This patch introduces a PCI DMA API and some generic code to support other DMA
APIs. It introduces a IOVector type that contains physical address/length
pairs. These vectors can be translated by the PCI layer and passed either to
generic copying functions or directly to the block or network subsy
uint32_t is the wrong type to use to represent physical addresses.
This patch is unchanged since v1.
Signed-off-by: Anthony Liguori <[EMAIL PROTECTED]>
diff --git a/cpu-all.h b/cpu-all.h
index 2a2b197..9e5d33b 100644
--- a/cpu-all.h
+++ b/cpu-all.h
@@ -834,7 +834,7 @@ typedef uint32_t CPUReadMem
This patch implements the virtio network driver backend. In KVM, this driver
can achieve 1gbit tx/rx performance. More patches are required to improve the
network IO infrastructure to achieve better performance in QEMU.
Since v1, I've updated the patch based on the IOVector refactoring.
Signed-
This patch implements the virtio balloon driver backend. A user can interact
with the balloon driver using a newly introduce monitor command 'balloon'.
Ballooning is used to request the guest to stop using a certain portion of its
memory. The guest notifies the host of this memory so the host ca
This patch introduces virtio support over PCI. virtio is a generic virtual IO
framework for Linux first introduced in 2.6.23. Since 2.6.25, virtio has
supported a PCI transport which this patch implements.
Since the last time these patches were posted to qemu-devel, I've reworked it
to use the p
This patch implements the virtio block driver backend.
Since v1, I've updated the patch based on the IOVector refactoring and fixed
an endianness bug in the config space.
Signed-off-by: Anthony Liguori <[EMAIL PROTECTED]>
diff --git a/Makefile.target b/Makefile.target
index 3ea40d1..f9fe660 1006
52 matches
Mail list logo
