Hi,
Le 16/03/2015 22:48, John Snow a écrit :
On 03/14/2015 12:50 PM, Hervé Poussineau wrote:
They were introduced in 6f7e9aec5eb5bdfa57a9e458e391b785c283a007 and
82407d1a4035e5bfefb53ffdcb270872f813b34c and lots of bug fixes were done after
that.
This fixes (at least) the detection of the flo
On 03/14/2015 12:50 PM, Hervé Poussineau wrote:
They were introduced in 6f7e9aec5eb5bdfa57a9e458e391b785c283a007 and
82407d1a4035e5bfefb53ffdcb270872f813b34c and lots of bug fixes were done after
that.
This fixes (at least) the detection of the floppy controller on Debian
4.0r9/SPARC,
and SS
On 03/16/2015 04:57 PM, Max Reitz wrote:
On 2015-03-13 at 14:30, John Snow wrote:
Add bdrv_clear_dirty_bitmap and a matching QMP command,
qmp_block_dirty_bitmap_clear that enables a user to reset
the bitmap attached to a drive.
This allows us to reset a bitmap in the event of a full
drive bac
On 2015-03-13 at 14:30, John Snow wrote:
Add bdrv_clear_dirty_bitmap and a matching QMP command,
qmp_block_dirty_bitmap_clear that enables a user to reset
the bitmap attached to a drive.
This allows us to reset a bitmap in the event of a full
drive backup.
Signed-off-by: John Snow
---
block.
On 2015-03-13 at 14:30, John Snow wrote:
Add the "frozen" status booleans, to inform clients
when a bitmap is occupied doing a task.
Signed-off-by: Fam Zheng
Signed-off-by: John Snow
---
block.c | 1 +
qapi/block-core.json | 5 -
2 files changed, 5 insertions(+), 1 deletio
On 2015-03-16 at 16:53, John Snow wrote:
On 03/16/2015 04:44 PM, Max Reitz wrote:
On 2015-03-13 at 14:30, John Snow wrote:
The new command pair is added to manage a user created dirty bitmap.
The
dirty bitmap's name is mandatory and must be unique for the same
device,
but different devices
On 2015-03-13 at 14:30, John Snow wrote:
We add a bitmap merge operation to assist in error cases
where we wish to combine two bitmaps together.
This is algorithmically O(bits) provided HBITMAP_LEVELS remains
constant. For a full bitmap on a 64bit machine:
sum(bits/64^k, k, 0, HBITMAP_LEVELS) ~=
On 2015-03-13 at 14:30, John Snow wrote:
As a convenience: between incremental backups, bitmap migrations
and bitmap persistence we seem to need to recalculate these a lot.
Because the lengths are a little bit-twiddly, let's just solidly
cache them and be done with it.
Signed-off-by: John Snow
On 03/16/2015 04:44 PM, Max Reitz wrote:
On 2015-03-13 at 14:30, John Snow wrote:
The new command pair is added to manage a user created dirty bitmap. The
dirty bitmap's name is mandatory and must be unique for the same device,
but different devices can have bitmaps with the same names.
The g
On 2015-03-13 at 14:30, John Snow wrote:
The new command pair is added to manage a user created dirty bitmap. The
dirty bitmap's name is mandatory and must be unique for the same device,
but different devices can have bitmaps with the same names.
The granularity is an optional field. If it is no
On 2015-03-13 at 14:30, John Snow wrote:
Signed-off-by: John Snow
---
docs/bitmaps.md | 311
1 file changed, 311 insertions(+)
create mode 100644 docs/bitmaps.md
Reviewed-by: Max Reitz
From: Peter Lieven
The VHD spec [1] allows for total_sectors of 65535 x 16 x 255 (~127GB)
represented by a CHS geometry. If total_sectors is greater
than 65535 x 16 x 255 this geometry is set as a maximum.
Qemu, Hyper-V and disk2vhd use this special geometry as an indicator
to use the image curr
The CHS calculation as done per the VHD spec imposes a maximum image
size of ~127 GB. Real VHD images exist that are larger than that.
Apparently there are two separate non-standard ways to achieve this:
You could use more heads than the spec does - this is the option that
qemu-img create chooses.
From: Fam Zheng
Signed-off-by: Fam Zheng
Message-id: 1425296209-1476-4-git-send-email-f...@redhat.com
Reviewed-by: Max Reitz
Signed-off-by: Max Reitz
---
blockdev.c | 92 --
1 file changed, 59 insertions(+), 33 deletions(-)
diff --g
From: Peter Lieven
*pnum can't be greater than s->block_size / BDRV_SECTOR_SIZE for allocated
sectors since there is always a bitmap in between.
Signed-off-by: Peter Lieven
Reviewed-by: Max Reitz
Message-id: 1425379316-19639-2-git-send-email...@kamp.de
Signed-off-by: Max Reitz
---
block/vpc.
From: Fam Zheng
Signed-off-by: Fam Zheng
Message-id: 1425296209-1476-2-git-send-email-f...@redhat.com
Reviewed-by: Max Reitz
Signed-off-by: Max Reitz
---
monitor.c | 9 +
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/monitor.c b/monitor.c
index c86a89e..8b703f9 100644
From: Max Reitz
When choosing a new place for the refcount table, alloc_refcount_block()
tries to infer the number of clusters used so far from its argument
cluster_index (which comes from the idea that if any cluster with an
index greater than cluster_index was in use, the refcount table would
h
From: Fam Zheng
All callers are converted, so drop it.
Signed-off-by: Fam Zheng
Message-id: 1425296209-1476-5-git-send-email-f...@redhat.com
Reviewed-by: Max Reitz
Signed-off-by: Max Reitz
---
block.c | 9 -
include/block/block.h | 1 -
2 files changed, 10 deletions(-)
From: Peter Lieven
The code to check the bitmap for the allocation status of each sector
has been "disabled by reason" ever since the vpc driver existed.
The reason might be that we might end up reading sector by sector
in vpc_read if we really used it. This would be a performance desaster.
The
From: Fam Zheng
Signed-off-by: Fam Zheng
Message-id: 1425296209-1476-3-git-send-email-f...@redhat.com
Reviewed-by: Max Reitz
Signed-off-by: Max Reitz
---
migration/block.c | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/migration/block.c b/migration/block.c
index 0c
From: Max Reitz
It is easy to create only self-referential refblocks, but there are
cases where that is impossible. This adds a test for two of those cases
(combined in a single test case).
Suggested-by: Eric Blake
Signed-off-by: Max Reitz
Message-id: 1417798412-15330-1-git-send-email-mre...@r
From: Max Reitz
Signed-off-by: Max Reitz
Message-id: 1423598552-24301-3-git-send-email-mre...@redhat.com
Reviewed-by: Eric Blake
Reviewed-by: Kevin Wolf
Signed-off-by: Max Reitz
---
tests/qemu-iotests/121 | 102 +
tests/qemu-iotests/121.out |
From: Markus Armbruster
qerror_report_err() is a transitional interface to help with
converting existing monitor commands to QMP. It should not be used
elsewhere. Replace by error_report_err().
Commit 6936f29 cleaned that up in qemu-img.c, but two calls have crept
in since. Take care of them
From: Peter Lieven
the field is named current size in the spec. Name it accordingly.
Signed-off-by: Peter Lieven
Reviewed-by: Max Reitz
Message-id: 1425379316-19639-5-git-send-email...@kamp.de
Signed-off-by: Max Reitz
---
block/vpc.c | 9 +
1 file changed, 5 insertions(+), 4 deletion
From: Markus Armbruster
Commit c4bacaf improved error reporting, but neglected to update
051.out. Commit 2726958 tried to redress, but didn't get it quite
right (punctuation difference), and shortly after commit
ae071cc..master improved error reporting some more, neglecting 051.out
some more. S
From: Markus Armbruster
We've steered users away from QCOW/QCOW2 encryption for a while,
because it's a flawed design (commit 136cd19 Describe flaws in
qcow/qcow2 encryption in the docs).
In addition to flawed crypto, we have comically bad usability, and
plain old bugs. Let me show you.
= Exam
From: Markus Armbruster
Message quality regressed in commit dc523cd.
Signed-off-by: Markus Armbruster
Signed-off-by: Kevin Wolf
---
qemu-img.c | 6 ++
tests/qemu-iotests/061.out | 1 -
2 files changed, 2 insertions(+), 5 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
i
From: Markus Armbruster
Error classes are a leftover from the days of "rich" error objects.
New code should always use ERROR_CLASS_GENERIC_ERROR. Commit e246211
added a use of ERROR_CLASS_DEVICE_NOT_FOUND. Replace it.
Signed-off-by: Markus Armbruster
Reviewed-by: Eric Blake
Signed-off-by: Ke
The following changes since commit dcf848c478dd8765bd4f746fc4e80eaad44cf87d:
Merge remote-tracking branch 'remotes/pmaydell/tags/pull-target-arm-20150316'
into staging (2015-03-16 13:56:10 +)
are available in the git repository at:
git://repo.or.cz/qemu/kevin.git tags/fo
On 03/16/2015 10:08 AM, Max Reitz wrote:
On 2015-03-13 at 07:09, Kevin Wolf wrote:
Am 13.03.2015 um 11:46 hat Peter Lieven geschrieben:
Hi,
I just stumbled across an old ticket where a user complains that an
ejected CD is visible again after a reset.
It seems that the behaviour of qemu chang
Am 12.03.2015 um 16:08 hat Markus Armbruster geschrieben:
> qerror_report_err() is a transitional interface to help with
> converting existing monitor commands to QMP. It should not be used
> elsewhere. Replace by error_report_err().
>
> Commit 6936f29 cleaned that up in qemu-img.c, but two call
Am 13.03.2015 um 18:51 hat Markus Armbruster geschrieben:
> Error classes are a leftover from the days of "rich" error objects.
> New code should always use ERROR_CLASS_GENERIC_ERROR. Commit e246211
> added a use of ERROR_CLASS_DEVICE_NOT_FOUND. Replace it.
>
> Signed-off-by: Markus Armbruster
On 2015-03-16 at 10:51, Paolo Bonzini wrote:
On 16/03/2015 15:49, Max Reitz wrote:
Right, but a better model than block.c should be for example
block/raw-posix.c, which returns ENOTSUP (I checked now...).
Maybe we should catch ENOTSUP in bdrv_co_discard() and override it?
That's what it does:
On 16/03/2015 15:48, Max Reitz wrote:
> On 2015-03-16 at 10:42, Paolo Bonzini wrote:
>>
>> On 16/03/2015 14:51, Max Reitz wrote:
Propagating the return value from write_sync is uglier, but it is even
better in terms of returned value.
>>> We can only return -errno values, but write_sync
On 16/03/2015 15:49, Max Reitz wrote:
>>>
>> Right, but a better model than block.c should be for example
>> block/raw-posix.c, which returns ENOTSUP (I checked now...).
>
> Maybe we should catch ENOTSUP in bdrv_co_discard() and override it?
That's what it does:
if (ret && ret != -ENOT
On 2015-03-16 at 10:44, Paolo Bonzini wrote:
On 16/03/2015 14:58, Max Reitz wrote:
diff --git a/block/nbd-client.c b/block/nbd-client.c
index be6803d..ab13607 100644
--- a/block/nbd-client.c
+++ b/block/nbd-client.c
@@ -315,6 +315,7 @@ int nbd_client_co_flush(BlockDriverState *bs)
ssize_
On 2015-03-16 at 10:42, Paolo Bonzini wrote:
On 16/03/2015 14:51, Max Reitz wrote:
Propagating the return value from write_sync is uglier, but it is even
better in terms of returned value.
We can only return -errno values, but write_sync() may do partial writes
so it may return non-negative va
On 16/03/2015 14:58, Max Reitz wrote:
>>>
>>> diff --git a/block/nbd-client.c b/block/nbd-client.c
>>> index be6803d..ab13607 100644
>>> --- a/block/nbd-client.c
>>> +++ b/block/nbd-client.c
>>> @@ -315,6 +315,7 @@ int nbd_client_co_flush(BlockDriverState *bs)
>>> ssize_t ret;
>>> i
On 16/03/2015 14:51, Max Reitz wrote:
>>>
>> Propagating the return value from write_sync is uglier, but it is even
>> better in terms of returned value.
>
> We can only return -errno values, but write_sync() may do partial writes
> so it may return non-negative values which still indicate an er
On 2015-03-13 at 07:09, Kevin Wolf wrote:
Am 13.03.2015 um 11:46 hat Peter Lieven geschrieben:
Hi,
I just stumbled across an old ticket where a user complains that an ejected CD
is visible again after a reset.
It seems that the behaviour of qemu changed somewhen in the past (maybe years
ago).
On 2015-03-11 at 07:31, Paolo Bonzini wrote:
On 25/02/2015 19:08, Max Reitz wrote:
If some operation cannot be performed by a block driver, it is normally
supposed to return an error. In these cases, however, it is fine to
pretend the operations were carried out successfully because if the NBD
On 2015-03-11 at 07:24, Paolo Bonzini wrote:
On 25/02/2015 19:08, Max Reitz wrote:
In case the connection is closed before the export length can be read,
and an export name had been specified, this generally indicates that for
some reason the export could not be opened (e.g. there is no export
On 2015-03-11 at 07:30, Paolo Bonzini wrote:
On 25/02/2015 19:08, Max Reitz wrote:
This patch makes sure the result of strtol() does not overflow (by
storing it in long integers instead of plain integers, and by checking
errno), allows the user to specify "--discard on" and
"--detect-zeroes unm
On 2015-03-11 at 07:22, Paolo Bonzini wrote:
On 25/02/2015 19:08, Max Reitz wrote:
Just returning -EINVAL for everything is bad. -EIO is often better, and
sometimes there is an even more fitting value.
Propagating the return value from write_sync is uglier, but it is even
better in terms of re
Am 13.03.2015 um 21:09 hat Markus Armbruster geschrieben:
> We've steered users away from QCOW/QCOW2 encryption for a while,
> because it's a flawed design (commit 136cd19 Describe flaws in
> qcow/qcow2 encryption in the docs).
>
> In addition to flawed crypto, we have comically bad usability, and
Am 14.03.2015 um 10:23 hat Markus Armbruster geschrieben:
> "check -T -qcow2" now passes again. Sorry for the mess I made.
>
> Markus Armbruster (2):
> iotests: Update 051's reference output
> qemu-img: Fix convert, amend error messages for unknown options
Thanks, applied to the block branch
On Fri, Mar 13, 2015 at 09:09:40PM +0100, Markus Armbruster wrote:
> We've steered users away from QCOW/QCOW2 encryption for a while,
> because it's a flawed design (commit 136cd19 Describe flaws in
> qcow/qcow2 encryption in the docs).
>
> In addition to flawed crypto, we have comically bad usabi
47 matches
Mail list logo