On 04/06/12 03:57, Rusty Russell wrote:
Unfortunately, this conflicts with Asias He's deadlock fix, which has
us just using the block-layer-supplied spinlock.
If we drop the lock around the kick as you suggest, we're playing with
fire. All the virtio backends have an atomic notify, so
Anthony Liguori anth...@codemonkey.ws writes:
On 05/29/2012 04:14 PM, Markus Armbruster wrote:
Luiz Capitulinolcapitul...@redhat.com writes:
On Mon, 28 May 2012 12:17:04 +0100
Stefan Hajnoczistefa...@linux.vnet.ibm.com wrote:
What we need to decide is whether it's okay to drop QEMU VLANs
On 06/04/2012 09:30 AM, Rob Landley wrote:
On 05/31/2012 12:46 PM, H. Peter Anvin wrote:
On 05/31/2012 01:01 AM, Raghavendra K T wrote:
+
+TODO:
+1. more information on input and output needed?
+2. Add more detail to purpose of hypercalls.
1. definitely, including the hypercall ABI.
On Mon, Jun 04, 2012 at 11:27:39AM +0930, Rusty Russell wrote:
On Wed, 30 May 2012 15:39:05 +0200, Christian Borntraeger
borntrae...@de.ibm.com wrote:
On 30/05/12 15:19, Stefan Hajnoczi wrote:
Holding the vblk-lock across kick causes poor scalability in SMP
guests. If one CPU is doing
On 06/01/2012 05:13 PM, Stefan Hajnoczi wrote:
Holding the vblk-lock across kick causes poor scalability in SMP
guests. If one CPU is doing virtqueue kick and another CPU touches the
vblk-lock it will have to spin until virtqueue kick completes.
This patch reduces system% CPU utilization in
On 06/04/2012 09:57 AM, Rusty Russell wrote:
On Wed, 30 May 2012 15:39:05 +0200, Christian
Borntraegerborntrae...@de.ibm.com wrote:
On 30/05/12 15:19, Stefan Hajnoczi wrote:
Holding the vblk-lock across kick causes poor scalability in SMP
guests. If one CPU is doing virtqueue kick and
On 05/31/2012 11:14 PM, Randy Dunlap wrote:
On 05/31/2012 01:01 AM, Raghavendra K T wrote:
From: Raghavendra K Traghavendra...@linux.vnet.ibm.com
Thanks Alex for KVM_HC_FEATURES inputs and Jan for VAPIC_POLL_IRQ
Signed-off-by: Raghavendra K Traghavendra...@linux.vnet.ibm.com
---
diff --git
On Sun, Jun 03, 2012 at 06:25:33PM +0200, Johannes Bauer wrote:
Therefore, I've uploaded the compressed trace.dat file, so you can maybe
have a look why the report tool barfs and interpret it correctly. I
can't figure it out. The trace is here:
http://spornkuller.de/trace.dat.bz2
I can
On 06/04/2012 09:57 AM, Rusty Russell wrote:
On Wed, 30 May 2012 15:39:05 +0200, Christian
Borntraegerborntrae...@de.ibm.com wrote:
On 30/05/12 15:19, Stefan Hajnoczi wrote:
Holding the vblk-lock across kick causes poor scalability in SMP
guests. If one CPU is doing virtqueue kick and
On 06/04/2012 11:53 AM, Gleb Natapov wrote:
On Sun, Jun 03, 2012 at 06:25:33PM +0200, Johannes Bauer wrote:
Therefore, I've uploaded the compressed trace.dat file, so you can maybe
have a look why the report tool barfs and interpret it correctly. I
can't figure it out. The trace is here:
Those constants will be also used by postcopy.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c |7 ---
arch_init.h |7 +++
2 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 38e0173..bd4e61e 100644
--- a/arch_init.c
This will be used by postcopy.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c |4 ++--
migration.h |2 ++
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index a9e8b74..38e0173 100644
--- a/arch_init.c
+++ b/arch_init.c
@@
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 13 -
arch_init.h |1 +
2 files changed, 9 insertions(+), 5 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 2617478..22d9691 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -203,6 +203,12 @@ int
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
qemu-file.h |1 +
savevm.c| 12
2 files changed, 13 insertions(+), 0 deletions(-)
diff --git a/qemu-file.h b/qemu-file.h
index 331ac8b..98a8023 100644
--- a/qemu-file.h
+++ b/qemu-file.h
@@ -71,6 +71,7 @@ QEMUFile
Later the structure will be shared.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
savevm.c | 14 +++---
1 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/savevm.c b/savevm.c
index 4b560b3..2fb0c3e 100644
--- a/savevm.c
+++ b/savevm.c
@@ -187,14 +187,14 @@
Introduce nonblocking fd read backend of QEMUFile.
This will be used by postcopy live migration.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
qemu-file.h |1 +
savevm.c| 40
2 files changed, 41 insertions(+), 0 deletions(-)
diff
Introduce a new method to drain the buffer of QEMUBufferedFile.
When postcopy migration, buffer size can increase unboundedly.
To keep the buffer size reasonably small, introduce the method to
wait for buffer to drain.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
buffered_file.c |
fd is already stored in QEMUFile so drop duplicated member
QEMUFileSocket::fd.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
savevm.c |4 +---
1 files changed, 1 insertions(+), 3 deletions(-)
diff --git a/savevm.c b/savevm.c
index ec9f5d0..4b560b3 100644
--- a/savevm.c
+++
This will be used by postcopy migration.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
migration.c |4 ++--
migration.h |2 ++
2 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/migration.c b/migration.c
index 753addb..48a8f68 100644
--- a/migration.c
+++
Those will be used by postcopy.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
qemu-file.h |3 +++
savevm.c|6 +++---
2 files changed, 6 insertions(+), 3 deletions(-)
diff --git a/qemu-file.h b/qemu-file.h
index 31b83f6..a285bef 100644
--- a/qemu-file.h
+++ b/qemu-file.h
This patch prepares for postcopy livemigration.
It introduces -postcopy option and its internal flag, migration_postcopy.
It introduces -postcopy-flags for chaging the behavior of incoming postcopy
mainly for benchmark/debug.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
migration.h
This patch introduces helper function for postcopy to access
umem char device and to communicate between incoming-qemu and umemd.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
changes v1 - v2:
- code simplification
- make fault trigger more robust
- introduce struct umem_pages
---
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
hmp-commands.hx |5 +++--
hmp.c|3 ++-
migration.c |8 +++-
migration.h |1 +
qapi-schema.json |2 +-
qmp-commands.hx |2 +-
savevm.c |1 +
7 files changed, 16 insertions(+), 6
This patch implements postcopy live migration for outgoing part
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
Changes v1 - v2:
- fix parameter to qemu_fdopen()
- handle QEMU_UMEM_REQ_EOC properly
when PO_STATE_ALL_PAGES_SENT, QEMU_UMEM_REQ_EOC request was ignored.
handle properly
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
hmp-commands.hx | 15 ++-
hmp.c|3 +++
migration.c | 20
migration.h |2 ++
qapi-schema.json |3 ++-
5 files changed, 37 insertions(+), 6 deletions(-)
diff --git
This is used by postcopy to tell the total length of QEMU_VM_SECTION_FULL
and QEMU_VM_SUBSECTION from outgoing to incoming.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
savevm.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/savevm.c b/savevm.c
index
When page is requested, send surrounding pages are also sent.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
migration-postcopy.c | 56 +
1 files changed, 51 insertions(+), 5 deletions(-)
diff --git a/migration-postcopy.c
Added -p option to migrate command for postcopy mode and
introduce postcopy parameter for migration to indicate that postcopy mode
is enabled.
Add -n option for postcopy migration which indicates disabling background
transfer.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
Chnages v1 -
When movebg mode is enabled, the point to send background page is set
to the next page to on-demand page.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
migration-postcopy.c |8
1 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/migration-postcopy.c
This patch implements postcopy live migration for incoming part
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
Changes v3 - v4:
- fork umemd early to address qemu devices touching guest ram via
post/pre_load
- code clean up on initialization
- Makefile.target
migration-postcopy.c is
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 21 -
exec.c | 12 ++--
2 files changed, 14 insertions(+), 19 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 9981abe..73bf250 100644
--- a/arch_init.c
+++ b/arch_init.c
@@
Cc: Andrea Arcangeli aarca...@redhat.com
Cc: Avi Kivity a...@redhat.com
Cc: Paolo Bonzini pbonz...@redhat.com
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
mm/memcontrol.c |1 +
mm/mempolicy.c |1 +
mm/shmem.c |1 +
3 files changed, 3 insertions(+), 0 deletions(-)
This is Linux kernel driver for qemu/kvm postcopy live migration.
This is used by qemu/kvm postcopy live migration patch.
TODO:
- Consider FUSE/CUSE option
So far several mmap patches for FUSE/CUSE are floating around. (their
purpose isn't different from our purpose, though). They haven't
This is a character device to hook page access.
The page fault in the area is propagated to another user process by
this chardriver. Then, the process fills the page contents and
resolves the page fault.
Cc: Andrea Arcangeli aarca...@redhat.com
Cc: Avi Kivity a...@redhat.com
Cc: Paolo Bonzini
This is used by postcopy live migration.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
buffered_file.c | 50 ++
buffered_file.h | 10 ++
2 files changed, 60 insertions(+), 0 deletions(-)
diff --git a/buffered_file.c
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
linux-headers/linux/umem.h | 42 ++
1 files changed, 42 insertions(+), 0 deletions(-)
create mode 100644 linux-headers/linux/umem.h
diff --git a/linux-headers/linux/umem.h
Add enable/disable postcopy mode. No dynamic test yet.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
configure | 12
1 files changed, 12 insertions(+), 0 deletions(-)
diff --git a/configure b/configure
index 1f338f8..21de4cb 100755
--- a/configure
+++ b/configure
@@
Introduce MigrationParams for parameters of migration.
Cc: Orit Wasserman owass...@redhat.com
Cc: Juan Quintela quint...@redhat.com
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
Changes v1 - v2:
- catch up qapi change
---
block-migration.c |8
migration.c | 21
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
migration.c |1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff --git a/migration.c b/migration.c
index 3f485d3..753addb 100644
--- a/migration.c
+++ b/migration.c
@@ -367,7 +367,6 @@ static MigrationState *migrate_init(int
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
buffered_file.c | 115 +++
buffered_file.h | 13 ++
2 files changed, 128 insertions(+), 0 deletions(-)
diff --git a/buffered_file.c b/buffered_file.c
index 22dd4c9..5198923 100644
Now qemu_file_fd() replaces qemu_stdio_fd().
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
migration-exec.c |4 ++--
migration-fd.c |2 +-
qemu-file.h |1 -
savevm.c | 12
4 files changed, 3 insertions(+), 16 deletions(-)
diff --git
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
scripts/update-linux-headers.sh |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/scripts/update-linux-headers.sh b/scripts/update-linux-headers.sh
index 9d2a4bc..2afdd54 100755
--- a/scripts/update-linux-headers.sh
This patch factors out buffer management logic.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
buffered_file.c | 141 +-
buffered_file.h |8 +++
2 files changed, 94 insertions(+), 55 deletions(-)
diff --git a/buffered_file.c
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
exec-obsolete.h |1 +
exec.c |4 ++--
2 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/exec-obsolete.h b/exec-obsolete.h
index 792c831..fb21dd7 100644
--- a/exec-obsolete.h
+++ b/exec-obsolete.h
@@ -25,6 +25,7
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 67 +-
arch_init.h |1 +
2 files changed, 39 insertions(+), 29 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index c861e30..bb0cd52 100644
--- a/arch_init.c
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 25 ++---
arch_init.h |7 +++
2 files changed, 25 insertions(+), 7 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 2a53f58..36ece1d 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -374,21
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
cpu-all.h |2 ++
exec.c| 51 +--
2 files changed, 31 insertions(+), 22 deletions(-)
diff --git a/cpu-all.h b/cpu-all.h
index 028528f..ff7f827 100644
--- a/cpu-all.h
+++ b/cpu-all.h
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 24
1 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 73bf250..2617478 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -155,8 +155,9 @@ static int
This will be used later by postcopy migration.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
qemu-file.h |1 +
savevm.c|5 +
2 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/qemu-file.h b/qemu-file.h
index a285bef..880ef4b 100644
--- a/qemu-file.h
+++
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 21 ++---
migration.h |1 +
2 files changed, 15 insertions(+), 7 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 36ece1d..28e5abb 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -287,6 +287,19 @@
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 28
migration.h |1 +
2 files changed, 17 insertions(+), 12 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index 900cc8e..c861e30 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -311,9
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c | 13 +
arch_init.h |1 +
2 files changed, 14 insertions(+), 0 deletions(-)
diff --git a/arch_init.c b/arch_init.c
index bb0cd52..9981abe 100644
--- a/arch_init.c
+++ b/arch_init.c
@@ -397,6 +397,19 @@ int
Introduce RAM_SAVE_VERSION_ID to represent version_id for ram save format.
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
arch_init.c |2 +-
arch_init.h |2 ++
vl.c|4 ++--
3 files changed, 5 insertions(+), 3 deletions(-)
diff --git a/arch_init.c b/arch_init.c
After the long time, we have v2. This is qemu part.
The linux kernel part is sent separatedly.
Changes v1 - v2:
- split up patches for review
- buffered file refactored
- many bug fixes
Espcially PV drivers can work with postcopy
- optimization/heuristic
Patches
1 - 30: refactoring exsiting
Signed-off-by: Isaku Yamahata yamah...@valinux.co.jp
---
Chnages v1 - v2:
- don't refer last_block which can be NULL.
And avoid possible infinite loop.
---
arch_init.c | 82 +-
arch_init.h |1 +
2 files changed, 48 insertions(+), 35
On Fri, Jun 01, 2012 at 10:13:06AM +0100, Stefan Hajnoczi wrote:
Holding the vblk-lock across kick causes poor scalability in SMP
guests. If one CPU is doing virtqueue kick and another CPU touches the
vblk-lock it will have to spin until virtqueue kick completes.
This patch reduces system%
On Fri, Jun 01, 2012 at 10:13:06AM +0100, Stefan Hajnoczi wrote:
Other block drivers (cciss, rbd, nbd) use spin_unlock_irq() so I followed
that.
To me this seems wrong: blk_run_queue() uses spin_lock_irqsave() but we enable
irqs with spin_unlock_irq(). If the caller of blk_run_queue() had
On Sun, 3 Jun 2012, Avi Kivity wrote:
On 06/01/2012 09:26 PM, Jan Kiszka wrote:
you suggesting we need a request_edge_threaded_only_irq() API? Thanks,
I'm just wondering if that restriction for threaded IRQs is really
necessary for all use cases we have. Threaded MSIs do not appear
On 2012-06-04 13:21, Thomas Gleixner wrote:
On Sun, 3 Jun 2012, Avi Kivity wrote:
On 06/01/2012 09:26 PM, Jan Kiszka wrote:
you suggesting we need a request_edge_threaded_only_irq() API? Thanks,
I'm just wondering if that restriction for threaded IRQs is really
necessary for all use
Hi,
I'm running lots of kvm instances for doing kernel boot tests.
Unfortunately the test system itself is not stable enough, I got scary
errors in both kvm and the host kernel. Like this.
[294025.795382] kvm used greatest stack depth: 2896 bytes left
[310388.622083] kvm[1864]: segfault at c ip
mmu_shrink() needlessly iterates over all VMs even though it will not
attempt to free mmu pages from more than one on them. Fix that and also
check used mmu pages count outside of VM lock to skip inactive VMs faster.
Signed-off-by: Gleb Natapov g...@redhat.com
diff --git a/arch/x86/kvm/mmu.c
On Mon, Jun 04, 2012 at 07:46:03PM +0800, Fengguang Wu wrote:
Hi,
I'm running lots of kvm instances for doing kernel boot tests.
Unfortunately the test system itself is not stable enough, I got scary
errors in both kvm and the host kernel. Like this.
What do you mean by in both kvm and
On 06/04/2012 05:57 PM, Isaku Yamahata wrote:
After the long time, we have v2. This is qemu part.
The linux kernel part is sent separatedly.
Changes v1 - v2:
- split up patches for review
- buffered file refactored
- many bug fixes
Espcially PV drivers can work with postcopy
-
On Mon, 4 Jun 2012, Jan Kiszka wrote:
On 2012-06-04 13:21, Thomas Gleixner wrote:
So this shortcut requires some checks before being applied to a specific
MSI/MSI-X vector.
Taking KVM aside, my general question remains if threaded MSI handlers
of all devices really need to apply
On Mon, 04 Jun 2012 12:56:41 +0800
Anthony Liguori anth...@codemonkey.ws wrote:
On 05/25/2012 08:53 PM, Luiz Capitulino wrote:
On Fri, 25 May 2012 13:01:37 +0100
Stefan Hajnoczistefa...@gmail.com wrote:
I agree it would be nice to drop entirely but I don't feel happy doing
that to
On Mon, Jun 04, 2012 at 05:01:30AM -0700, Chegu Vinod wrote:
Hello Isaku Yamahata,
Hi.
I just saw your patches..Would it be possible to email me a tar bundle of
these
patches (makes it easier to apply the patches to a copy of the upstream
qemu.git)
I uploaded them to github for those who
On 2012-06-04 15:07, Thomas Gleixner wrote:
On Mon, 4 Jun 2012, Jan Kiszka wrote:
On 2012-06-04 13:21, Thomas Gleixner wrote:
So this shortcut requires some checks before being applied to a specific
MSI/MSI-X vector.
Taking KVM aside, my general question remains if threaded MSI handlers
of
On Mon, 4 Jun 2012, Jan Kiszka wrote:
On 2012-06-04 15:07, Thomas Gleixner wrote:
On Mon, 4 Jun 2012, Jan Kiszka wrote:
On 2012-06-04 13:21, Thomas Gleixner wrote:
So this shortcut requires some checks before being applied to a specific
MSI/MSI-X vector.
Taking KVM aside, my
Hi list,
I have recently started using KVM, and I'm looking to see how I might
be able to use and adapt KVM for my own purposes, to do things more
easily on some of the virtual machines I use.
To this end, I was wondering if it is possible to execute a script or
run a command or a process within
On Mon, Jun 04, 2012 at 08:37:04PM +0800, Anthony Liguori wrote:
On 06/04/2012 05:57 PM, Isaku Yamahata wrote:
After the long time, we have v2. This is qemu part.
The linux kernel part is sent separatedly.
Changes v1 - v2:
- split up patches for review
- buffered file refactored
- many
The following changes since commit 75b4a94628c5035b5c350eff4a4edf923b5dc33a:
pci-assign: Drop kvm_assigned_irq::host_irq initialization (2012-06-01
20:51:10 -0300)
are available in the git repository at:
git://git.kiszka.org/qemu-kvm.git queues/qemu-merge
This is the back-merge of post-1.1
Am 01.06.2012 10:31, schrieb Kevin Wolf:
Am 01.06.2012 09:57, schrieb Ren, Yongjie:
-Original Message-
From: Marcelo Tosatti [mailto:mtosa...@redhat.com]
Sent: Thursday, May 31, 2012 4:28 AM
To: Ren, Yongjie
Cc: Kevin Wolf; Avi Kivity; kvm@vger.kernel.org; Liu, RongrongX
Subject:
On 6/4/2012 6:13 AM, Isaku Yamahata wrote:
On Mon, Jun 04, 2012 at 05:01:30AM -0700, Chegu Vinod wrote:
Hello Isaku Yamahata,
Hi.
I just saw your patches..Would it be possible to email me a tar bundle of these
patches (makes it easier to apply the patches to a copy of the upstream
qemu.git)
https://bugzilla.kernel.org/show_bug.cgi?id=43339
Summary: Wrong Pci-Bridge Header Type check.
Product: Virtualization
Version: unspecified
Kernel Version: 3.4
Platform: All
OS/Version: Linux
Tree: Mainline
Status:
The queue size for virtio_blk is 256 and AIO_MAX is 32, we might be
short of available aio events if guest issues 32 requests
simultaneously. Following error is observed when guest running stressed
I/O workload.
Info: disk_image__read error: total=-11
To fix this, let's increase the aio
Keep trying if io_submit returns EAGAIN. No need to fail the request.
Signed-off-by: Asias He asias.he...@gmail.com
---
tools/kvm/util/read-write.c | 16 +---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/tools/kvm/util/read-write.c b/tools/kvm/util/read-write.c
On Mon, Jun 04, 2012 at 07:27:25AM -0700, Chegu Vinod wrote:
On 6/4/2012 6:13 AM, Isaku Yamahata wrote:
On Mon, Jun 04, 2012 at 05:01:30AM -0700, Chegu Vinod wrote:
Hello Isaku Yamahata,
Hi.
I just saw your patches..Would it be possible to email me a tar bundle of
these
patches (makes it
We read and write in sectors by default. It makes little sense to add
the extra _sector string for read and write ops/function name.
Signed-off-by: Asias He asias.he...@gmail.com
---
tools/kvm/disk/blk.c |4 ++--
tools/kvm/disk/core.c |8
All blk requests are processed in notify_vq() which is in the context of
ioeventfd thread: ioeventfd__thread(). The processing in notify_vq() may
take a long time to complete and all devices share the single ioeventfd
thead, so this might block other device's notify_vq() being called and
starve
On Mon, Jun 04, 2012 at 11:40:53PM +0800, Asias He wrote:
+static void *virtio_blk_thread(void *dev)
+{
+ struct blk_dev *bdev = dev;
+ u64 data;
+
+ while (1) {
+ read(bdev-io_efd, data, sizeof(u64));
+ virtio_blk_do_io(bdev-kvm, bdev-vqs[0], bdev);
On Mon, 2012-06-04 at 23:40 +0800, Asias He wrote:
All blk requests are processed in notify_vq() which is in the context of
ioeventfd thread: ioeventfd__thread(). The processing in notify_vq() may
take a long time to complete and all devices share the single ioeventfd
thead, so this might
On Mon, Jun 04, 2012 at 08:35:30PM +0800, Fengguang Wu wrote:
Hi Gleb,
On Mon, Jun 04, 2012 at 02:56:50PM +0300, Gleb Natapov wrote:
On Mon, Jun 04, 2012 at 07:46:03PM +0800, Fengguang Wu wrote:
Hi,
I'm running lots of kvm instances for doing kernel boot tests.
Unfortunately the
On 04.06.2012 10:53, Gleb Natapov wrote:
On Sun, Jun 03, 2012 at 06:25:33PM +0200, Johannes Bauer wrote:
Therefore, I've uploaded the compressed trace.dat file, so you can maybe
have a look why the report tool barfs and interpret it correctly. I
can't figure it out. The trace is here:
On 04.06.2012 20:28, Johannes Bauer wrote:
What is your command line?
bin/qemu-system-x86_64 -cpu host -enable-kvm -net nic -net
user,smb=Share,restrict=on -drive
media=disk,file=Windows7_x32.qcow2,if=virtio -m 2048 -smp 1 -nographic
Just noticed that the output I just provided was for the
On Mon, Jun 4, 2012 at 11:48 PM, Cyrill Gorcunov gorcu...@openvz.org wrote:
On Mon, Jun 04, 2012 at 11:40:53PM +0800, Asias He wrote:
+static void *virtio_blk_thread(void *dev)
+{
+ struct blk_dev *bdev = dev;
+ u64 data;
+
+ while (1) {
+ read(bdev-io_efd, data,
Am 04.06.2012 07:46, schrieb Anthony Liguori:
On 05/22/2012 12:37 AM, Avi Kivity wrote:
Please pull from:
git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
Pulled. Thanks.
This broke the ppc build. Guys, why wasn't this tested? There's only
three KVM targets to test compared
On Tue, Jun 5, 2012 at 12:07 AM, Sasha Levin levinsasha...@gmail.com wrote:
On Mon, 2012-06-04 at 23:40 +0800, Asias He wrote:
All blk requests are processed in notify_vq() which is in the context of
ioeventfd thread: ioeventfd__thread(). The processing in notify_vq() may
take a long time to
Allows to use cpu_reset() in place of cpu_state_reset().
Signed-off-by: Andreas Färber afaer...@suse.de
Reviewed-by: Igor Mammedov imamm...@redhat.com
---
cpu-exec.c |4 ++--
target-i386/cpu.h|4 ++--
target-i386/helper.c | 13 -
target-i386/kvm.c|6
On Thu, 2012-05-24 at 09:02 -0300, Jan Kiszka wrote:
Since PCI 2.3, this bit is mandatory, and it should be independent of
the masking bit. The question is, if your device is supposed to support
2.3, thus is just buggy,
It's a PCI Express device :-)
or if our detection algorithm is
Yep, that's what I'd suggest as well, add a blacklist to
pci_intx_mask_supported() so this device returns false and we require an
exclusive interrupt for it. Thanks,
BTW, we should consider supporting an MSI-only option for guests as
well:
LSIs are a problem for virtualization, especially
On 05.06.2012, at 03:39, Benjamin Herrenschmidt wrote:
Yep, that's what I'd suggest as well, add a blacklist to
pci_intx_mask_supported() so this device returns false and we require an
exclusive interrupt for it. Thanks,
BTW, we should consider supporting an MSI-only option for guests
On 06/05/2012 08:52 AM, Andreas Färber wrote:
Am 04.06.2012 07:46, schrieb Anthony Liguori:
On 05/22/2012 12:37 AM, Avi Kivity wrote:
Please pull from:
git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
Pulled. Thanks.
This broke the ppc build. Guys, why wasn't this tested?
Am 05.06.2012 03:58, schrieb Anthony Liguori:
Is build bot running against uq/master? If it's not, maybe we should
add it to build bot to catch this sort of thing.
That's a question for Stefan and Daniel to answer.
Regards,
Andreas
--
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg,
On Tue, 2012-06-05 at 03:44 +0200, Alexander Graf wrote:
Wouldn't the no LSI setting be a flag to the vfio configuration? So
when you set up the device group, you say this group can only do
MSI. That way the interface would be sysfs and QEMU wouldn't have
anything to do with it :)
Sure
On Mon, 2012-06-04 at 16:11 -0500, Chris Sanders wrote:
Hello, I've been working for several days trying to get Pci
Passthrough to work. So far the #kvm IRC channel has helped me with a
few suggestions, though that hasn't yet solved the problem. I'm
running CentOS 6.2 and was suggested I try
Jan:
sorry for late response of your suggestion.
I have found the patch which produce this problem, it comes from
this one: 7850ac5420803996e2960d15b924021f28e0dffc.
I change as the following, it works fine.
diff -ur -i kvm-kmod-3.4/x86/kvm_main.c kvm-kmod-3.4-fix/x86/kvm_main.c
Am 04.06.2012 07:46, schrieb Anthony Liguori:
On 05/22/2012 12:37 AM, Avi Kivity wrote:
Please pull from:
git://git.kernel.org/pub/scm/virt/kvm/qemu-kvm.git uq/master
Pulled. Thanks.
This broke the ppc build. Guys, why wasn't this tested? There's only
three KVM targets to test compared
Am 05.06.2012 03:58, schrieb Anthony Liguori:
Is build bot running against uq/master? If it's not, maybe we should
add it to build bot to catch this sort of thing.
That's a question for Stefan and Daniel to answer.
Regards,
Andreas
--
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg,
98 matches
Mail list logo