27.08.2019 23:12, John Snow wrote:
>
>
> On 8/23/19 5:22 AM, Vladimir Sementsov-Ogievskiy wrote:
>> 14.08.2019 13:07, Vladimir Sementsov-Ogievskiy wrote:
>>> To get rid of implicit filters related workarounds in future let's
>>> deprecate them now.
>>
>> Interesting, could we deprecate implicit
On Fri, Aug 23, 2019 at 03:49:24PM -0400, Dmitry Fomichev wrote:
> +uint8_t bdrv_is_zoned(BlockDriverState *bs)
> +{
> +/*
> + * Host Aware zone devices are supposed to be able to work
> + * just like regular block devices. Thus, we only consider
> + * Host Managed devices to be
On Thu, Aug 22, 2019 at 11:34:24AM -0700, Raphael Norwitz wrote:
> Same rational as: e6cc11d64fc998c11a4dfcde8fda3fc33a74d844
>
> Of the 3 virtqueues, seabios only sets cmd, leaving ctrl
> and event without a physical address. This can cause
> vhost_verify_ring_part_mapping to return ENOMEM,
On Tue, 2019-08-27 at 18:22 -0400, John Snow wrote:
> Without a commit message, I have no real hope of reviewing this. I was
> CC'd, though, so I'll give it a blind shot.
>
> We want to add write_zeroes support for block/nvme, but I can't really
> verify any of that is correct or working without
On Tue, 2019-08-27 at 18:29 -0400, John Snow wrote:
>
> On 8/25/19 3:15 AM, Maxim Levitsky wrote:
> > Signed-off-by: Maxim Levitsky
> > ---
> > block/nvme.c | 83 ++
> > block/trace-events | 2 ++
> > 2 files changed, 85 insertions(+)
> >
> >
23.08.2019 17:34, Eric Blake wrote:
> While it may be counterintuitive at first, the introduction of
> NBD_CMD_WRITE_ZEROES and NBD_CMD_BLOCK_STATUS has caused a performance
> regression in qemu [1], when copying a sparse file. When the
> destination file must contain the same contents as the
Nir Soffer wrote:
> Replace confusing usage:
>
> ~BDRV_SECTOR_MASK
>
> With more clear:
>
> (BDRV_SECTOR_SIZE - 1)
>
> Remove BDRV_SECTOR_MASK and the unused BDRV_BLOCK_OFFSET_MASK which was
> it's last user.
>
> Signed-off-by: Nir Soffer
Reviewed-by: Juan Quintela
On Fri, Aug 23, 2019 at 03:49:27PM -0400, Dmitry Fomichev wrote:
> diff --git a/block/file-posix.c b/block/file-posix.c
> index d9f2fc5e46..090e7c4d2f 100644
> --- a/block/file-posix.c
> +++ b/block/file-posix.c
> @@ -2884,6 +2884,20 @@ static int raw_check_perm(BlockDriverState *bs,
> uint64_t
Patchew URL:
https://patchew.org/QEMU/20190828125654.10544-1-dplotni...@virtuozzo.com/
Hi,
This series seems to have some coding style problems. See output below for
more information:
Message-id: 20190828125654.10544-1-dplotni...@virtuozzo.com
Type: series
Subject: [Qemu-devel] [PATCH v4
In our documentation, we use a mix of "$QEMU", "qemu-system-i386" and
"qemu-system-x86_64" when we give examples to the users how to run
QEMU. Some more consistency would be good here. Also some distributions
use different names for the QEMU binary (e.g. "qemu-kvm" in RHEL), so
providing more
On Fri, Aug 23, 2019 at 03:49:26PM -0400, Dmitry Fomichev wrote:
> Added a new boolean argument to blkconf_apply_backend_options()
> to let the common block code know whether the chosen block
> backend can handle zoned block devices or not.
>
> blkconf_apply_backend_options() then sets
The patch allow to process image compression type defined
in the image header and choose an appropriate method for
image clusters (de)compression.
Signed-off-by: Denis Plotnikov
---
block/qcow2-threads.c | 78 +++
1 file changed, 64 insertions(+), 14
On Fri, Aug 23, 2019 at 03:03:39PM +0200, Max Reitz wrote:
> Hi,
>
> As suggested by Paolo, this series drops xfsctl() calls where we have
> working fallocate() alternatives. (And thus replaces “block/file-posix:
> Fix xfs_write_zeroes()”.)
>
> Unfortunately, we also use xfsctl() to inquire the
On 8/28/19 4:57 AM, Vladimir Sementsov-Ogievskiy wrote:
>> Hence, it is desirable to have a way for clients to specify that a
>> particular write zero request is being attempted for a fast wipe, and
>> get an immediate failure if the zero request would otherwise take the
>> same time as a write.
The patch adds some preparation parts for incompatible compression type
feature to QCOW2 header that indicates that *all* compressed clusters
must be (de)compressed using a certain compression type.
It is implied that the compression type is set on the image creation and
can be changed only later
v4:
* remove not feasible switch case [Vladimir]
* add sanity checks to zstd decompresssion [Vladimir]
* store zstd compressed length in big endian [Max, Kevin]
v3:
* relax the compression type setting requirement when
the compression type is not zlib [Eric, Kevin]
* add compression type values
On Fri, Aug 23, 2019 at 03:49:23PM -0400, Dmitry Fomichev wrote:
> Dmitry Fomichev (4):
> block: Add zoned device model property
> raw: Recognize zoned backing devices
> block/ide/scsi: Set BLK_PERM_SUPPORT_ZONED
> raw: Don't open ZBDs if backend can't handle them
The overall approach
On Fri, Aug 23, 2019 at 03:49:25PM -0400, Dmitry Fomichev wrote:
> +static int hdev_get_zoned_model(int fd)
Please use the enum:
static BdrvZonedModel hdev_get_zoned_model(int fd)
> +{
> +#ifdef CONFIG_LINUX
> +char buf[32];
> +int ret;
> +
> +ret = hdev_read_blk_queue_entry(fd,
zstd significantly reduces cluster compression time.
It provides better compression performance maintaining
the same level of compression ratio in comparison with
zlib, which, at the moment, has been the only compression
method available.
The performance test results:
Test compresses and
On 27/08/2019 22:42, John Snow wrote:
>
>
> On 8/23/19 11:27 AM, Vladimir Sementsov-Ogievskiy wrote:
>> 16.08.2019 4:01, John Snow wrote:
>>>
>>>
>>> On 7/19/19 12:30 PM, Andrey Shinkevich wrote:
To synchronize the time when QEMU is running longer under the Valgrind,
increase the
28.08.2019 16:04, Eric Blake wrote:
> On 8/28/19 4:57 AM, Vladimir Sementsov-Ogievskiy wrote:
>
>>> Hence, it is desirable to have a way for clients to specify that a
>>> particular write zero request is being attempted for a fast wipe, and
>>> get an immediate failure if the zero request would
On 28.08.19 16:27, Vladimir Sementsov-Ogievskiy wrote:
> 28.08.2019 17:22, Max Reitz wrote:
>> On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
>>> Split copying logic which will be shared with backup-top filter.
>>>
>>> Signed-off-by: Vladimir Sementsov-Ogievskiy
>>> ---
>>>
On 27/08/2019 22:56, John Snow wrote:
>
>
> On 8/25/19 11:26 AM, Andrey Shinkevich wrote:
>>
>>
>> On 16/08/2019 01:49, John Snow wrote:
>>>
>>>
>>> On 7/19/19 12:30 PM, Andrey Shinkevich wrote:
With the '-valgrind' option, let all the QEMU processes be run under
the Valgrind tool.
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Fix comment style and reflow arguments in same manner like
> block_copy_state_new.
I like the current function header style better.
Max
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
> block/backup.c | 40
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Split block_copy to separate file, to be cleanly shared with backup-top
> filter driver in further commits.
>
> It's a clean movement, the only change is drop "static" from interface
> functions.
>
> Signed-off-by: Vladimir
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Split copying code part from backup to "block-copy", including separate
> state structure and function renaming. This is needed to share it with
> backup-top filter driver in further commits.
>
> Notes:
>
> 1. As BlockCopyState keeps own
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Teach bdrv_debug_breakpoint and bdrv_debug_remove_breakpoint skip
> filters with backing. This is needed to implement and use in backup job
> it's own backup_top filter driver (like mirror already has one), and
> without this improvement,
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> We shouldn't try to copy bytes beyond EOF. Fix it.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
> block/backup.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/block/backup.c b/block/backup.c
> index
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Split copying logic which will be shared with backup-top filter.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
> block/backup.c | 47 ---
> 1 file changed, 32 insertions(+), 15
On 27/08/2019 22:45, John Snow wrote:
>
>
> On 8/25/19 11:24 AM, Andrey Shinkevich wrote:
>>
>>
>> On 16/08/2019 03:55, John Snow wrote:
>>>
>>>
>>> On 7/19/19 12:30 PM, Andrey Shinkevich wrote:
The Valgrind uses the exported variable TMPDIR and fails if the
directory does not exist.
28.08.2019 17:22, Max Reitz wrote:
> On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
>> Split copying logic which will be shared with backup-top filter.
>>
>> Signed-off-by: Vladimir Sementsov-Ogievskiy
>> ---
>> block/backup.c | 47 ---
>> 1
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> After backup-top filter appearing it's not possible to see dirty
> bitmaps in top node, so use node-name instead.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
> tests/qemu-iotests/124| 3 +-
> tests/qemu-iotests/257
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> After previous commit Drive.device is actually unused. Drop it together
> with .name property. While being here reuse .node in qmp commands
> instead of writing 'drive0' twice.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
>
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Backup-top filter caches write operations and does copy-before-write
> operations.
>
> The driver will be used in backup instead of write-notifiers.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
> block/backup-top.h | 37
(Peter: search for "pkrempa" down below.)
On 8/28/19 5:20 AM, Vladimir Sementsov-Ogievskiy wrote:
> 27.08.2019 23:12, John Snow wrote:
>>
>>
>> On 8/23/19 5:22 AM, Vladimir Sementsov-Ogievskiy wrote:
>>> 14.08.2019 13:07, Vladimir Sementsov-Ogievskiy wrote:
To get rid of implicit filters
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> From: Vladimir Sementsov-Ogievskiy
Hm. :-)
Do you want to fix that?
> Introduce lock/unlock range functionality, based on serialized
> requests. This is needed to refactor backup, dropping local
> tracked-request-like synchronization.
>
On 8/28/19 11:24 AM, Andrey Shinkevich wrote:
>
>
> On 27/08/2019 22:42, John Snow wrote:
>>
>>
>> On 8/23/19 11:27 AM, Vladimir Sementsov-Ogievskiy wrote:
>>> 16.08.2019 4:01, John Snow wrote:
On 7/19/19 12:30 PM, Andrey Shinkevich wrote:
> To synchronize the time when
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> SCSI devices are unused in test, drop them.
>
> Signed-off-by: Vladimir Sementsov-Ogievskiy
> ---
> tests/qemu-iotests/257 | 8 ---
> tests/qemu-iotests/257.out | 44 --
> 2 files changed, 52
On 8/26/19 11:50 AM, Andrey Shinkevich wrote:
> The Valgrind tool fails to manage its termination in multi-threaded
> processes when they raise the signal SIGKILL. The bug has been reported
> to the Valgrind maintainers and was registered as the bug #409141:
>
On 8/26/19 11:50 AM, Andrey Shinkevich wrote:
> To synchronize the time when QEMU is running longer under the Valgrind,
> increase the sleeping time in the test 247.
>
> Signed-off-by: Andrey Shinkevich
> Reviewed-by: Vladimir Sementsov-Ogievskiy
Probably fine, as discussed. We can work on
On 8/26/19 11:50 AM, Andrey Shinkevich wrote:
> With the '-valgrind' option, let all the QEMU processes be run under
> the Valgrind tool. The Valgrind own parameters may be set with its
> environment variable VALGRIND_OPTS, e.g.
> $ VALGRIND_OPTS="--leak-check=yes" ./check -valgrind
> or they
On 8/28/19 5:34 AM, Thomas Huth wrote:
> In our documentation, we use a mix of "$QEMU", "qemu-system-i386" and
> "qemu-system-x86_64" when we give examples to the users how to run
> QEMU. Some more consistency would be good here. Also some distributions
> use different names for the QEMU binary
On 8/27/19 2:59 PM, Nir Soffer wrote:
> While working on 4k support, I noticed that there is lot of code using
> BDRV_SECTOR_SIZE (512) for checking alignment. I wonder how this can work with
> 4k storage.
>
> Lets start by cleaning up to make the code easier to understand:
> - Use
On 8/27/19 2:59 PM, Nir Soffer wrote:
> Replace confusing usage:
>
> ~BDRV_SECTOR_MASK
>
> With more clear:
>
> (BDRV_SECTOR_SIZE - 1)
>
> Remove BDRV_SECTOR_MASK and the unused BDRV_BLOCK_OFFSET_MASK which was
> it's last user.
>
Kind of lateral in my opinion, but if there was
On 8/23/19 9:03 AM, Max Reitz wrote:
> This exercises the regression introduced in commit
> 50ba5b2d994853b38fed10e0841b119da0f8b8e5. On my machine, it has close
> to a 50 % false-negative rate, but that should still be sufficient to
> test the fix.
>
> Signed-off-by: Max Reitz
> ---
>
On 8/23/19 9:03 AM, Max Reitz wrote:
> This patch removes xfs_write_zeroes() and xfs_discard(). Both functions
> have been added just before the same feature was present through
> fallocate():
>
> - fallocate() has supported PUNCH_HOLE for XFS since Linux 2.6.38 (March
> 2011);
On 26.08.19 18:13, Vladimir Sementsov-Ogievskiy wrote:
> Drop write notifiers and use filter node instead.
>
> = Changes =
>
> 1. add filter-node-name argument for backup qmp api. We have to do it
> in this commit, as 257 needs to be fixed.
I feel a bit bad about it not being an implicit node.
On 28/08/2019 21.18, John Snow wrote:
>
>
> On 8/28/19 5:34 AM, Thomas Huth wrote:
>> In our documentation, we use a mix of "$QEMU", "qemu-system-i386" and
>> "qemu-system-x86_64" when we give examples to the users how to run
>> QEMU. Some more consistency would be good here. Also some
On 8/28/19 5:58 PM, John Snow wrote:
>> +++ b/tests/qemu-iotests/common.rc
>> @@ -60,61 +60,132 @@ if ! . ./common.config
>> exit 1
>> fi
>>
>> +# Unset the variables to turn Valgrind off for specific processes, e.g.
That's not unsetting, that's setting to the empty string.
>> +# $
On 8/23/19 10:32 AM, Vladimir Sementsov-Ogievskiy wrote:
> In job_finish_sync job_enter should be enough for a job to make some
> progress and draining is a wrong tool for it. So use job_enter directly
> here and drop job_drain with all related staff not used more.
>
> Suggested-by: Kevin Wolf
Gentle ping. This should be fairly easy to review, I hope; the worst of
it is making sure that no tests remain that don't engage an entry point
in iotests.py anymore.
On 8/20/19 7:52 PM, John Snow wrote:
> This series uses python logging to enable output conditionally on
> iotests.log(). We unify
51 matches
Mail list logo
