From: Vladimir Sementsov-Ogievskiy
qmp_block_dirty_bitmap_add and do_block_dirty_bitmap_remove do acquire
aio context since 0a6c86d024c52b. But this is not enough: we also must
lock qcow2 mutex when access in-image metadata. Especially it concerns
freeing qcow2 clusters.
To achieve this, move qc
From: Vladimir Sementsov-Ogievskiy
mutex field is just a pointer to bs->dirty_bitmap_mutex, so no needs
to store it in BdrvDirtyBitmap when we have bs pointer in it (since
previous patch).
Drop mutex field. Constantly use bdrv_dirty_bitmaps_lock/unlock in
block/dirty-bitmap.c to make it more obv
From: Vladimir Sementsov-Ogievskiy
block/dirty-bitmap.c seems to be more appropriate for it and
bdrv_remove_persistent_dirty_bitmap already in it.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: John Snow
Message-id: 20190920082543.23444-2-vsement...@virtuozzo.com
Signed-off-by: John
From: Vladimir Sementsov-Ogievskiy
It's more comfortable to not deal with local_err.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: John Snow
Message-id: 20190920082543.23444-3-vsement...@virtuozzo.com
Signed-off-by: John Snow
---
block/qcow2.h| 5 ++---
include/bl
From: Vladimir Sementsov-Ogievskiy
It's needed to fix reopening qcow2 with bitmaps to RW. Currently it
can't work, as qcow2 needs write access to file child, to mark bitmaps
in-image with IN_USE flag. But usually children goes after parents in
reopen queue and file child is still RO on qcow2 reop
From: Vladimir Sementsov-Ogievskiy
bdrv_dirty_bitmap_next is always used in same pattern. So, split it
into _next and _first, instead of combining two functions into one and
add FOR_EACH_DIRTY_BITMAP macro.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: John Snow
Message-id: 20190916
From: Vladimir Sementsov-Ogievskiy
Reopening bitmaps to RW was broken prior to previous commit. Check that
it works now.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Message-id: 20190927122355.7344-4-vsement...@virtuozzo.com
Signed-off-by: John Snow
---
tests/qemu-iotests/165 | 57
From: Vladimir Sementsov-Ogievskiy
Add bs field to BdrvDirtyBitmap structure. Drop BlockDriverState
parameter from bitmap APIs where possible.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: John Snow
Message-id: 20190916141911.5255-3-vsement...@virtuozzo.com
[Rebased on top of block-
From: Vladimir Sementsov-Ogievskiy
We'll need reverse-foreach in the following commit, QTAILQ support it,
so move to QTAILQ.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Max Reitz
Message-id: 20190927122355.7344-2-vsement...@virtuozzo.com
Signed-off-by: John Snow
---
include/bloc
On Thu, Oct 17, 2019 at 6:50 AM Peter Maydell wrote:
>
> Switch the xilinx_timer code away from bottom-half based ptimers to
> the new transaction-based ptimer API. This just requires adding
> begin/commit calls around the various places that modify the ptimer
> state, and using the new ptimer_in
From: Vladimir Sementsov-Ogievskiy
Signed-off-by: Vladimir Sementsov-Ogievskiy
Message-id: 20190927122355.7344-8-vsement...@virtuozzo.com
[Maintainer edit: removed 260 from auto group per Peter Maydell. --js]
Signed-off-by: John Snow
---
tests/qemu-iotests/260 | 89
From: Vladimir Sementsov-Ogievskiy
The function is unused, drop it.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: John Snow
Message-id: 20190927122355.7344-6-vsement...@virtuozzo.com
Signed-off-by: John Snow
---
block/qcow2.h| 2 --
block/qcow2-bitmap.c | 15 +
From: Vladimir Sementsov-Ogievskiy
Firstly, no reason to optimize failure path. Then, function name is
ambiguous: it checks for readonly and similar things, but someone may
think that it will ignore normal bitmaps which was just unchanged, and
this is in bad relation with the fact that we should
From: Vladimir Sementsov-Ogievskiy
- Correct check for write access to file child, and in correct place
(only if we want to write).
- Support reopen rw -> rw (which will be used in following commit),
for example, !bdrv_dirty_bitmap_readonly() is not a corruption if
bitmap is marked IN_USE i
From: Vladimir Sementsov-Ogievskiy
qcow2_reopen_bitmaps_ro wants to store bitmaps and then mark them all
readonly. But the latter don't work, as
qcow2_store_persistent_dirty_bitmaps removes bitmaps after storing.
It's OK for inactivation but bad idea for reopen-ro. And this leads to
the following
On Thu, Oct 17, 2019 at 6:53 AM Peter Maydell wrote:
>
> Switch the xilinx_axidma code away from bottom-half based ptimers to
> the new transaction-based ptimer API. This just requires adding
> begin/commit calls around the various places that modify the ptimer
> state, and using the new ptimer_i
From: Vladimir Sementsov-Ogievskiy
The only reason I can imagine for this strange code at the very-end of
bdrv_reopen_commit is the fact that bs->read_only updated after
calling drv->bdrv_reopen_commit in bdrv_reopen_commit. And in the same
time, prior to previous commit, qcow2_reopen_bitmaps_rw
On Thu, Oct 17, 2019 at 6:42 AM Peter Maydell wrote:
>
> Switch the fsl_etsec code away from bottom-half based ptimers to
> the new transaction-based ptimer API. This just requires adding
> begin/commit calls around the various places that modify the ptimer
> state, and using the new ptimer_init(
This parameter has been deprecated since 2.12.0 and is eligible for
removal. Remove this parameter as it is actually completely ignored;
let's not give false hope.
Signed-off-by: John Snow
Reviewed-by: Eric Blake
Reviewed-by: Vladimir Sementsov-Ogievskiy
Message-id: 20191002232411.29968-1-js...
On Wed, Oct 16, 2019 at 07:41:24PM -0300, Eduardo Habkost wrote:
> On Wed, Oct 16, 2019 at 08:11:57AM +0200, Thomas Huth wrote:
> > On 16/10/2019 05.00, Eduardo Habkost wrote:
> > > On Tue, Sep 17, 2019 at 08:31:40PM -0300, Eduardo Habkost wrote:
> > >> On Mon, Jul 01, 2019 at 07:25:27PM -0300, Edu
Quoting Philippe Mathieu-Daudé (2019-10-09 09:23:22)
> On 10/9/19 4:17 PM, Michael Roth wrote:
> > Quoting Philippe Mathieu-Daudé (2019-10-08 08:04:52)
> >> Hi Michael,
> >>
> >> On 10/2/19 1:44 AM, Michael Roth wrote:
> >>> Hi everyone,
> >>>
> >>> The following new patches are queued for QEMU sta
Hi everyone,
I am pleased to announce that the QEMU v4.0.1 stable release is now
available:
You can grab the tarball from our download page here:
https://www.qemu.org/download/#source
v4.0.1 is now tagged in the official qemu.git repository,
and the stable-4.0 branch has been updated accordin
On 10/17/19 5:29 PM, Eduardo Habkost wrote:
> On Thu, Oct 17, 2019 at 05:11:29PM -0400, John Snow wrote:
>>
>>
>> On 10/16/19 3:24 PM, Eduardo Habkost wrote:
>>> This series ports image-fuzzer to Python 3.
>>>
>>> Eduardo Habkost (10):
>>> image-fuzzer: Open image files in binary mode
>>> im
On Thu, Oct 17, 2019 at 07:05:41PM -0300, Eduardo Habkost wrote:
> On Wed, Oct 16, 2019 at 07:41:24PM -0300, Eduardo Habkost wrote:
> > On Wed, Oct 16, 2019 at 08:11:57AM +0200, Thomas Huth wrote:
> > > On 16/10/2019 05.00, Eduardo Habkost wrote:
> > > > On Tue, Sep 17, 2019 at 08:31:40PM -0300, Ed
We will soon implement the SYS_timer. This timer is used by Linux
in the thermal subsystem, so once available, the subsystem will be
enabled and poll the temperature sensors. We need to provide the
minimum required to keep Linux booting.
Add a dummy thermal sensor returning ~25°C based on:
https:/
From: PhilMD
Hi,
Some patches from v1 are already merged. This v2 addresses the
review comment from v1, and add patches to clean the memory
space when using multiple cores.
Laurent, if you test U-Boot with this patchset again, do you mind
replying with a "Tested-by:" tag?
The next patchset is
Add the 64-bit free running timer. Do not model the COMPARE register
(no IRQ generated).
This timer is used by Linux kernel and recently U-Boot:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/clocksource/bcm2835_timer.c?h=v3.7
https://github.com/u-boot/u-boot/blob/v
write_secondary_boot() is used in SMP configurations where the
CPU address space might not be the main System Bus.
The rom_add_blob_fixed_as() function allow us to specify an
address space. Use it to write each boot blob in the corresponding
CPU address space.
Signed-off-by: Philippe Mathieu-Daudé
Map the thermal sensor in the BCM2835 block.
Reviewed-by: Alistair Francis
Signed-off-by: Philippe Mathieu-Daudé
---
hw/arm/bcm2835_peripherals.c | 13 +
include/hw/arm/bcm2835_peripherals.h | 2 ++
include/hw/arm/raspi_platform.h | 1 +
3 files changed, 16 insertions
This file creates the BCM2836/BCM2837 blocks.
The biggest differences with the BCM2838 we are going to add, are
the base addresses of the interrupt controller and the peripherals.
Add these addresses in the BCM283XInfo structure to make this
block more modular. Remove the MCORE_OFFSET offset as it
Our code currently create the raspi2 (based on the BCM2836)
and the raspi3 (on the BCM2837). Similarly, the raspi4 is
based on the BCM2838. To be able to add the new board,
make the current code more modular:
- Dynamically fills the 'board-rev' value
- Allow DRAM sizes different than 1 GiB
Rename
write_secondary_boot() is used in SMP configurations where the
CPU address space might not be the main System Bus.
The rom_add_blob_fixed_as() function allow us to specify an
address space. Use it to write each boot blob in the corresponding
CPU address space.
Signed-off-by: Philippe Mathieu-Daudé
Connect the recently added SYS_timer.
Now U-Boot does not hang anymore polling a free running counter
stuck at 0.
This timer is also used by the Linux kernel thermal subsystem.
Signed-off-by: Philippe Mathieu-Daudé
---
v2: Remove spurious error check (Alex)
---
hw/arm/bcm2835_peripherals.c
Currently the VideoCore is created in the Peripheral container
as the 'GPU bus'. It is created there because the peripherals
using DMA use physical addresses from the VideoCore bus.
However the VideoCore is a GPU core placed at the same
hierarchical level than the ARM cores.
To match the datasheet
Currently all CPUs access the main system bus. Let each CPU have
his own address space.
Before:
address-space: cpu-secure-memory-0
- (prio 0, i/o): system
-3fff (prio 0, ram): ram
3f00-3fff (prio
On Thu, Oct 17, 2019 at 9:47 AM Peter Maydell wrote:
>
> On Thu, 17 Oct 2019 at 13:39, wrote:
> >
> > From: Carlos Santos
> >
> > uClibc defines _SC_LEVEL1_ICACHE_LINESIZE and _SC_LEVEL1_DCACHE_LINESIZE
> > but the corresponding sysconf calls returns -1, which is a valid result,
> > meaning that
On Thu, Oct 17, 2019 at 4:05 PM Philippe Mathieu-Daudé wrote:
>
> Connect the recently added SYS_timer.
> Now U-Boot does not hang anymore polling a free running counter
> stuck at 0.
> This timer is also used by the Linux kernel thermal subsystem.
>
> Signed-off-by: Philippe Mathieu-Daudé
Revie
In any of these `*_for_each_*` functions, the last entry in the buffer (so the
"remaining length in the buffer" `len` is equal to the length of the
entry `nlmsg_len`/`nla_len`/etc) has size that is not a multiple of the
alignment, the aligned lengths `*_ALIGN(*_len)` will be greater than `len`.
Sin
For the first target page, all_zero is set to true for this round check.
After target_pages introduced, we could leverage this variable instead
of checking the address offset.
Signed-off-by: Wei Yang
---
migration/ram.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/migrati
In this case, page_buffer content would not be used.
Skip this to save some time.
Signed-off-by: Wei Yang
---
migration/ram.c | 8 +++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/migration/ram.c b/migration/ram.c
index 99a98b2da4..7938a643d9 100644
--- a/migration/ram.c
++
Compress is not supported with postcopy, it is safe to wait for
decompress thread just in precopy.
This is a preparation for later patch.
Signed-off-by: Wei Yang
---
migration/ram.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/migration/ram.c b/migration/ram.c
index 793
In postcopy, it requires to place whole host page instead of target
page.
Currently, it relies on the page offset to decide whether this is the
last target page. We also can count the target page number during the
iteration. When the number of target page equals
(host page size / target page size)
This patch set tries enable compress during postcopy.
postcopy requires to place a whole host page, while migration thread migrate
memory in target page size. This makes postcopy need to collect all target
pages in one host page before placing via userfaultfd.
To enable compress during postcopy,
postcopy requires to place a whole host page, while migration thread
migrate memory in target page size. This makes postcopy need to collect
all target pages in one host page before placing via userfaultfd.
To enable compress during postcopy, there are two problems to solve:
1. Random order f
After using number of target page received to track one host page, we
could have the capability to handle random order target page arrival in
one host page.
This is a preparation for enabling compress during postcopy.
Signed-off-by: Wei Yang
---
migration/ram.c | 16 +++-
1 file cha
Hi Paolo,
Ping :)
On 10/11/2019 3:49 PM, Tao Xu wrote:
On 10/11/2019 3:41 PM, Xu, Tao3 wrote:
[...]
diff --git a/target/i386/kvm.c b/target/i386/kvm.c
index 11b9c854b5..a465c893b5 100644
--- a/target/i386/kvm.c
+++ b/target/i386/kvm.c
@@ -401,6 +401,12 @@ uint32_t kvm_arch_get_supported_cpuid(KV
On Fri, 18 Oct 2019 at 01:17, Felipe Franciosi wrote:
>
> Originally, vhost-user-scsi did not implement a handle_output callback
> as that didn't seem necessary. Turns out it is.
>
> Depending on which other devices are presented to a VM, SeaBIOS may
> decide to map vhost-user-scsi devices on the
Sergio Lopez writes:
> Markus Armbruster writes:
>
>> Philippe Mathieu-Daudé writes:
>>
>>> Hi Sergio,
>>>
>>> On 10/15/19 1:23 PM, Sergio Lopez wrote:
Follow checkpatch.pl recommendation and replace the use of strtol with
qemu_strtol in x86_load_linux().
>>>
>>> "with qemu_strtoui"
>
Cornelia Huck writes:
> On Thu, 17 Oct 2019 16:40:56 +0200
> Thomas Huth wrote:
>
>> On 17/10/2019 16.34, Cornelia Huck wrote:
>> > On Thu, 17 Oct 2019 16:21:23 +0200
>> > Thomas Huth wrote:
>> >
>> >> There is no USB on s390x, so running qemu-system-s390x with
>> >> "-machine ...,usb=on" is
On 17/10/2019 16:42, Cédric Le Goater wrote:
> The interrupt presenters are not reseted today.
I should have added :
... when CPU are hot-plugged.
> Extend the sPAPR IRQ
> backend with a new cpu_intc_reset() handler which will be called by
> the CPU reset handler.
>
> spapr_realize_vcpu() is m
On 17/10/2019 18.41, Peter Maydell wrote:
> On Fri, 27 Sep 2019 at 17:44, Max Reitz wrote:
>>
>> On 27.09.19 18:39, Peter Maydell wrote:
>>> Hi; I just saw this iotest failure (on an s390x box, as it happens):
>>>
>>> TESTiotest-qcow2: 130 [fail]
>>> QEMU --
>>> "/home/linux1/qemu/b
On 17/10/2019 20.18, Philippe Mathieu-Daudé wrote:
> On 10/17/19 4:40 PM, Thomas Huth wrote:
>> On 17/10/2019 16.34, Cornelia Huck wrote:
>>> On Thu, 17 Oct 2019 16:21:23 +0200
>>> Thomas Huth wrote:
>>>
There is no USB on s390x, so running qemu-system-s390x with
"-machine ...,usb=on" is
On 17/10/2019 18.18, Thomas Huth wrote:
> On 17/10/2019 18.07, Stefan Hajnoczi wrote:
>> On Thu, Oct 17, 2019 at 04:52:54PM +0200, Thomas Huth wrote:
>>> On 11/10/2019 10.56, Stefan Hajnoczi wrote:
Implement the VIRTIO 1.0 virtio-pci interface. The main change here is
that the register l
On 18/10/2019 08.48, Thomas Huth wrote:
> On 17/10/2019 18.18, Thomas Huth wrote:
>> On 17/10/2019 18.07, Stefan Hajnoczi wrote:
>>> On Thu, Oct 17, 2019 at 04:52:54PM +0200, Thomas Huth wrote:
On 11/10/2019 10.56, Stefan Hajnoczi wrote:
> Implement the VIRTIO 1.0 virtio-pci interface. Th
401 - 454 of 454 matches
Mail list logo
