On Wed, Jul 26, 2017 at 04:12:21PM +0100, Stefan Hajnoczi wrote:
On Wed, Jul 26, 2017 at 05:19:24PM +0300, Manos Pitsidianakis wrote:
This proposal follows a discussion we had with Kevin and Stefan on filter
node management.
With block filter drivers arises a need to configure filter nodes
. This is done by gathering ThrottleGroup membership
details from BlockBackendPublic into ThrottleGroupMember and refactoring
existing code to use the structure.
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blo
{
"iops-total" : 99
}
}
}
This also means a group's configuration can be fetched with qom-get.
ThrottleGroups can be anonymous which means they can't get accessed by
other users ie they will always be units instead of group (Because they
have one ThrottleGroupM
Move the CoMutex and CoQueue inits inside throttle_group_register_tgm()
which is called whenever a ThrottleGroupMember is initialized. There's
no need for them to be separate.
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/block-backend.c | 14 -
block/throttle-groups.c | 38 -
include/block/throttle-groups.h | 7 -
tests/test-throttle.c | 63 +---
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
tests/qemu-iotests/184 | 237 +
tests/qemu-iotests/184.out | 319 +
tests/qemu-iotests/group | 1 +
3 files changed, 557 insertions(+)
creat
for uint on error
move frees in throttle_group_obj_finalize()
split throttle_group_{set,get}()
add throttle_recurse_is_first_non_filter()
Manos Pitsidianakis (7):
block: move ThrottleGroup membership to ThrottleGroupMember
block: add aio_context field in ThrottleGroupMember
block: tidy
-obj-y += throttle.o
block-obj-y += crypto.o
diff --git a/block/throttle.c b/block/throttle.c
new file mode 100644
index 00..f3395462fb
--- /dev/null
+++ b/block/throttle.c
@@ -0,0 +1,395 @@
+/*
+ * QEMU block throttling filter driver infrastructure
+ *
+ * Copyright (c) 2017 Manos Pitsidi
This is needed to configure throttle filter driver nodes with QAPI.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
qapi/block-core.json | 20 +++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/qapi/block-core.json b/qapi/block-core.json
This proposal follows a discussion we had with Kevin and Stefan on
filter node management.
With block filter drivers arises a need to configure filter nodes on
runtime with QMP on live graphs. A problem with doing live graph
modifications is that some block jobs modify the graph when they are
On Tue, Jul 25, 2017 at 05:09:41PM +0100, Stefan Hajnoczi wrote:
On Tue, Jul 25, 2017 at 01:29:08PM +0300, Manos Pitsidianakis wrote:
On Mon, Jul 24, 2017 at 04:12:47PM +0100, Stefan Hajnoczi wrote:
> On Fri, Jul 14, 2017 at 12:45:18PM +0300, Manos Pitsidianakis wrote:
> > Thro
On Mon, Jul 24, 2017 at 04:12:47PM +0100, Stefan Hajnoczi wrote:
On Fri, Jul 14, 2017 at 12:45:18PM +0300, Manos Pitsidianakis wrote:
ThrottleGroup is converted to an object. This will allow the future
throttle block filter drive easy creation and configuration of throttle
groups in QMP and cli
ting README that ./check runs just the raw
format but it might be a good idea to mention that it's the default in
the Howto as well. For this patch though:
Reviewed-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
While we're on the topic of qemu-iotests, it'd be a nice thing to
parallelize the tests with a Makefile.
signature.asc
Description: PGP signature
patch is
not in the list archives either).
Reviewed-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
signature.asc
Description: PGP signature
On Fri, Jul 14, 2017 at 12:45:19PM +0300, Manos Pitsidianakis wrote:
+static BlockDriver bdrv_throttle = {
+.format_name= "throttle",
+.protocol_name = "throttle",
+.instance_size = sizeof
On Fri, Jul 14, 2017 at 09:42:22AM -0500, Eric Blake wrote:
On 07/14/2017 09:35 AM, Manos Pitsidianakis wrote:
In some error paths it is possible to QDECREF a freed dangling
explicit_options, resulting in a heap overflow crash. For example
bdrv_open_inherit()'s fail unrefs it, then calls
If open succeeds but we fail after, return the error and let callers
unref and delete their bs, while cleaning up their allocations.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 24
1 file changed, 12 insertions(+), 12 deletions(-)
diff -
In some error paths it is possible to QDECREF a freed dangling
explicit_options, resulting in a heap overflow crash. For example
bdrv_open_inherit()'s fail unrefs it, then calls bdrv_unref which calls
bdrv_close which also unrefs it.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua
v3:
new commit: block: fix dangling bs->explicit_options in block.c
rework error paths in bdrv_open_driver()
v2:
move bdrv_unref_child(bs, bs->file) to bdrv_open_driver
do not set bs->drv to NULL if open succeeds
Manos Pitsidianakis (2):
block: fix dangling bs->explicit_option
On Fri, Jul 14, 2017 at 03:15:06PM +0200, Pradeep Jagadeesh wrote:
Hi Manos,
Thanks for sharing the link to your code patch.
On 7/14/2017 2:22 PM, Manos Pitsidianakis wrote:
Hello Pradeep, you might be interested in my work on refactoring the
block layer's throttling interface in my series
Hello Pradeep, you might be interested in my work on refactoring the
block layer's throttling interface in my series:
https://lists.nongnu.org/archive/html/qemu-devel/2017-07/msg04191.html
In this series you copy the existing legacy interface we want to get rid
of. I think it will be easier
{
"iops-total" : 99
}
}
}
This also means a group's configuration can be fetched with qom-get.
ThrottleGroups can be anonymous which means they can't get accessed by
other users ie they will always be units instead of group (Because they
have one ThrottleGroupMe
-obj-y += throttle.o
block-obj-y += crypto.o
diff --git a/block/throttle.c b/block/throttle.c
new file mode 100644
index 00..5665ccd8b5
--- /dev/null
+++ b/block/throttle.c
@@ -0,0 +1,345 @@
+/*
+ * QEMU block throttling filter driver infrastructure
+ *
+ * Copyright (c) 2017 Manos Pitsidi
. This is done by gathering ThrottleGroup membership
details from BlockBackendPublic into ThrottleGroupMember and refactoring
existing code to use the structure.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
Notes:
Dropped R-b because I changed some of the comments (remov
timer_cb() needs to know about the current Aio context of the throttle
request that is woken up. In order to make ThrottleGroupMember backend
agnostic, this information is stored in an aio_context field instead of
accessing it from BlockBackend.
Signed-off-by: Manos Pitsidianakis <e
configurations (throttling at any point in the graph, chained filters)
There's also the question of how filter drivers affect the rest of the block
layer, especially live block jobs. This won't be tackled in this series though.
Manos Pitsidianakis (7):
block: move ThrottleGroup membership
This is needed to configure throttle filter driver nodes with QAPI.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
qapi/block-core.json | 20 +++-
1 file changed, 19 insertions(+), 1 deletion(-)
diff --git a/qapi/block-core.json b/qapi/block-core.json
Move the CoMutex and CoQueue inits inside throttle_group_register_tgm()
which is called whenever a ThrottleGroupMember is initialized. There's
no need for them to be separate.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/block-backend.c | 3 ---
block/throttle-gr
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
tests/qemu-iotests/184 | 237 +
tests/qemu-iotests/184.out | 319 +
tests/qemu-iotests/group | 1 +
3 files changed, 557 insertions(+)
creat
d-by: Stefan Hajnoczi <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 8 +---
1 file changed, 1 insertion(+), 7 deletions(-)
diff --git a/block/blkdebug.c b/block/blkdebug.c
index c19ab28f07..91ffd1fe12 100644
--- a/bloc
i <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 14 --
block/raw-format.c| 6 --
include/block/block.h | 1 -
include/block/block_int.h | 1 -
4 files changed, 22 deletions(-)
diff --git a/
octl()
new patch: remove duplicate code from block/raw-format.c
Manos Pitsidianakis (4):
block: pass bdrv_* methods to bs->file by default in block filters
block: remove unused bdrv_media_changed
block: remove bdrv_truncate callback in blkdebug
block: add default implementations for bdr
edhat.com>
Reviewed-by: Eric Blake <ebl...@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Reviewed-by: Kevin Wolf <kw...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 12 +---
block/commit.c
edhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 21 +++--
include/block/block_int.h | 6 +-
2 files changed, 24 insertions(+), 3 deletions(-)
diff --git a/block.c b/block.c
index d24ae85868..dd409bfb5b 100644
--- a/b
On Thu, Jul 13, 2017 at 06:23:14AM -0500, Eric Blake wrote:
On 07/13/2017 05:01 AM, Manos Pitsidianakis wrote:
The following functions fail if bs->drv is a filter and does not
implement them:
bdrv_probe_blocksizes
bdrv_probe_geometry
bdrv_truncate
bdrv_has_zero_init
bdrv_get_info
Inst
Now that bdrv_truncate is passed to bs->file by default, remove the
callback from block/blkdebug.c and set is_filter to true.
Reviewed-by: Eric Blake <ebl...@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.n
i <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 14 --
block/raw-format.c| 6 --
include/block/block.h | 1 -
include/block/block_int.h | 1 -
4 files changed, 22 deletions(-)
diff --git a/
edhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 21 +++--
include/block/block_int.h | 6 +-
2 files changed, 24 insertions(+), 3 deletions(-)
diff --git a/block.c b/block.c
index d24ae85868..892f29d231 100644
--- a/b
edhat.com>
Reviewed-by: Eric Blake <ebl...@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Reviewed-by: Kevin Wolf <kw...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.c
remove bdrv_truncate method from blkdebug
v2:
do not pass to bs->file if bs->drv is NULL
move bs->file check outside of bdrv_inc_in_flight() area in bdrv_co_ioctl()
new patch: remove duplicate code from block/raw-format.c
Manos Pitsidianakis (4):
block: pass bdrv_* methods to bs-&
On Tue, Jul 11, 2017 at 03:49:50PM +0100, Daniel P. Berrange wrote:
On Tue, Jul 11, 2017 at 04:44:46PM +0200, Markus Armbruster wrote:
Markus Armbruster <arm...@redhat.com> writes:
> Manos Pitsidianakis <el13...@mail.ntua.gr> writes:
>
>> Is there a specific reason th
On Wed, Jul 12, 2017 at 05:15:01PM +0200, Kevin Wolf wrote:
Am 12.07.2017 um 09:41 hat Markus Armbruster geschrieben:
Eric Blake <ebl...@redhat.com> writes:
> On 07/11/2017 11:37 AM, Manos Pitsidianakis wrote:
>> This function is not used anywhere, so remove it.
>>
>
On Wed, Jul 12, 2017 at 10:33:37AM +0200, Kevin Wolf wrote:
Am 11.07.2017 um 20:50 hat Manos Pitsidianakis geschrieben:
On Tue, Jul 11, 2017 at 05:16:17PM +0200, Kevin Wolf wrote:
>Am 01.07.2017 um 17:39 hat Manos Pitsidianakis geschrieben:
>>bdrv_open_driver() is called in t
On Wed, Jul 12, 2017 at 09:49:20AM +0200, Markus Armbruster wrote:
Manos Pitsidianakis <el13...@mail.ntua.gr> writes:
This series makes implementing some of the bdrv_* callbacks easier for block
filters by passing requests to bs->file if bs->drv doesn't implement it instea
On Tue, Jul 11, 2017 at 05:16:17PM +0200, Kevin Wolf wrote:
Am 01.07.2017 um 17:39 hat Manos Pitsidianakis geschrieben:
bdrv_open_driver() is called in two places, bdrv_new_open_driver() and
bdrv_open_common(). In the latter, failure cleanup in is in its caller,
bdrv_open_inherit(), which
ithout implementing them. This
commit makes `drv->is_filter = true` imply that these callbacks will be
forwarded to bs->file by default, so disabling support for these
functions must be done explicitly.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c
edhat.com>
Reviewed-by: Eric Blake <ebl...@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Reviewed-by: Kevin Wolf <kw...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 12 +---
block/commit.c
Now that bdrv_truncate is passed to bs->file by default, remove the
callback from block/blkdebug.c and set is_filter to true.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 7 +--
1 file changed, 1 insertion(+), 6 deletions(-)
diff --git a/block/b
estion
new patch: remove bdrv_truncate method from blkdebug
v2:
do not pass to bs->file if bs->drv is NULL
move bs->file check outside of bdrv_inc_in_flight() area in bdrv_co_ioctl()
new patch: remove duplicate code from block/raw-format.c
Manos Pitsidianakis (4):
block: pass bdrv_
This function is not used anywhere, so remove it.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 14 --
block/raw-format.c| 6 --
include/block/block.h | 1 -
include/block/block_int.h | 1 -
4 files changed, 22 del
Is there a specific reason this patch wasn't finished? If I'm not wrong
using non-scalar properties with -object is still not possible, yet
would be a very useful feature for drivers with UserCreatable objects.
Archive link since this is an old patch:
On Fri, Jul 07, 2017 at 11:28:15AM +0200, Kevin Wolf wrote:
Am 29.06.2017 um 22:06 hat Manos Pitsidianakis geschrieben:
On Thu, Jun 29, 2017 at 03:57:49PM +0200, Kevin Wolf wrote:
>Am 29.06.2017 um 14:07 hat Manos Pitsidianakis geschrieben:
>>On Thu, Jun 29, 2017 at 01:18:24PM +0200, K
Now that bdrv_truncate is passed to bs->file by default, remove the
callback from block/blkdebug.c
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 6 --
1 file changed, 6 deletions(-)
diff --git a/block/blkdebug.c b/block/blkdebug.c
index b25856c4.
Now that passing the call to bs->file is the default for some bdrv_*
callbacks, remove the duplicate implementations in block/raw-format.c
Reviewed-by: Eric Blake <ebl...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/ra
duplicate code from block/raw-format.c
Manos Pitsidianakis (4):
block: Pass bdrv_* methods to bs->file by default
block: Use defaults of bdrv_* callbacks in raw
block: Remove bdrv_truncate callback in blkdebug
block: Add default implementations for bdrv_co_get_block_status()
block.c
edhat.com>
Reviewed-by: Eric Blake <ebl...@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 12 +---
block/commit.c| 12 +---
filter drivers to support those methods without implementing them.
Reviewed-by: Eric Blake <ebl...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c| 27 +--
block/io.c | 5 +
2 files changed, 30 insertions(+), 2 deleti
On Mon, Jul 03, 2017 at 12:51:45PM -0500, Eric Blake wrote:
On 07/03/2017 11:12 AM, Eric Blake wrote:
On 06/29/2017 01:43 PM, Manos Pitsidianakis wrote:
bdrv_co_get_block_status_from_file() and
bdrv_co_get_block_status_from_backing() set *file to bs->file and
bs->backing respective
Since in the future ThrottleGroup configuration will be done even if it
has no members, we need to separate the two.
Manos Pitsidianakis (2):
block: add clock_type field to ThrottleGroup
block: remove timer canceling in throttle_config()
block/throttle-groups.c | 27
-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/throttle-groups.c | 20 ++--
fsdev/qemu-fsdev-throttle.c | 2 +-
include/qemu/throttle.h | 1 +
tests/test-throttle.c | 4 ++--
util/throttle.c | 4 +++-
5 files changed, 17 insertions(
requests will follow the new configuration. This
also allows a throttle group's configuration to be changed even when it
has no members.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/throttle-groups.c | 10 +-
fsdev/qemu-fsdev-throttle.c | 2 +-
includ
e care of all
callers and do not set bs->drv to NULL unless the driver's open function
failed. When bs is destroyed by removing its last reference, bdrv_close()
checks bs->drv to perform the needed cleanups and also call the driver's close
function.
Signed-off-by: Manos Pitsidianakis <el13.
On Thu, Jun 29, 2017 at 03:57:49PM +0200, Kevin Wolf wrote:
Am 29.06.2017 um 14:07 hat Manos Pitsidianakis geschrieben:
On Thu, Jun 29, 2017 at 01:18:24PM +0200, Kevin Wolf wrote:
>Am 29.06.2017 um 08:03 hat Manos Pitsidianakis geschrieben:
>>bdrv_open_driver() is called in t
<stefa...@redhat.com>
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/blkdebug.c | 12 +---
block/commit.c| 12 +---
block/io.c| 24
block/mirror.c| 12 +---
include/bloc
filter drivers to support those methods without implementing them.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c| 27 +--
block/io.c | 5 +
2 files changed, 30 insertions(+), 2 deletions(-)
diff --git a/block.c b/block.c
index 694396
Now that passing the call to bs->file is the default for some bdrv_*
callbacks, remove the duplicate implementations in block/raw-format.c
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/raw-format.c | 32 +---
1 file changed, 1 inserti
commit
0dd2ec6bc46bd93ea0b9df602962883417f31400
v2:
do not pass to bs->file if bs->drv is NULL
move bs->file check outside of bdrv_inc_in_flight() area in bdrv_co_ioctl()
new patch: remove duplicate code from block/raw-format.c
Manos Pitsidianakis (3):
block: pass bdrv_* methods to bs-&
On Thu, Jun 29, 2017 at 01:18:24PM +0200, Kevin Wolf wrote:
Am 29.06.2017 um 08:03 hat Manos Pitsidianakis geschrieben:
bdrv_open_driver() is called in two places, bdrv_new_open_driver() and
bdrv_open_common(). In the latter, failure cleanup in is in its caller,
bdrv_open_inherit(), which
off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/block.c b/block.c
index 694396281b..aeacd520e0 100644
--- a/block.c
+++ b/block.c
@@ -1165,6 +1165,9 @@ BlockDriverState *bdrv_new_open_driver(BlockDriver *drv,
const char *
On Wed, Jun 28, 2017 at 05:36:54PM +0200, Kevin Wolf wrote:
Am 28.06.2017 um 17:22 hat Manos Pitsidianakis geschrieben:
Since we're moving groups to QOM we will need ids for each group.
Can objects be anonymous?
Hm, that's a good question. But object_new() doesn't take an ID, so I
think
On Wed, Jun 28, 2017 at 04:40:12PM +0200, Kevin Wolf wrote:
Am 23.06.2017 um 14:46 hat Manos Pitsidianakis geschrieben:
block/throttle.c uses existing I/O throttle infrastructure inside a
block filter driver. I/O operations are intercepted in the filter's
read/write coroutines, and referred
On Wed, Jun 28, 2017 at 03:42:41PM +0200, Alberto Garcia wrote:
On Fri 23 Jun 2017 02:46:58 PM CEST, Manos Pitsidianakis wrote:
BlockDriverState *bdrv_new_open_driver(BlockDriver *drv, const char *node_name,
- int flags, Error **errp
On Wed, Jun 28, 2017 at 03:35:48PM +0200, Alberto Garcia wrote:
On Fri 23 Jun 2017 02:46:57 PM CEST, Manos Pitsidianakis wrote:
+# @BlockdevOptionsThrottle:
+#
+# Driver specific block device options for Throttle
+#
I would put this earlier in the json file, together with the rest
On Wed, Jun 28, 2017 at 01:27:36PM +0200, Kevin Wolf wrote:
Am 23.06.2017 um 14:46 hat Manos Pitsidianakis geschrieben:
timer_cb() needs to know about the current Aio context of the throttle
request that is woken up. In order to make ThrottleGroupMember backend
agnostic, this information
tors_inter < intermediate->total_sectors)) {
+(intermediate == top || sector_num + psectors_inter < size_inter))
{
n = psectors_inter;
}
--
2.9.4
Reviewed-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
signature.asc
Description: PGP signature
bdrv_co_get_block_status_from_file() and
bdrv_co_get_block_status_from_backing() set *file to bs->file and
bs->backing respectively, so that bdrv_co_get_block_status() can recurse
to them. Future block drivers won't have to duplicate code to implement
this.
Signed-off-by: Manos Pitsidi
filter drivers to support those methods without implementing them.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block.c| 45 ++---
block/io.c | 4
2 files changed, 38 insertions(+), 11 deletions(-)
diff --git a/block.c b/b
commit
0dd2ec6bc46bd93ea0b9df602962883417f31400
Manos Pitsidianakis (2):
block: pass bdrv_* methods to bs->file by default
block: add default implementations for bdrv_co_get_block_status()
block.c | 45 ++---
block/blkdebug.c
On Tue, Jun 27, 2017 at 06:05:55PM +0200, Alberto Garcia wrote:
On Mon 26 Jun 2017 06:58:32 PM CEST, Manos Pitsidianakis wrote:
On Mon, Jun 26, 2017 at 03:52:34PM +0100, Stefan Hajnoczi wrote:
On Fri, Jun 23, 2017 at 03:46:56PM +0300, Manos Pitsidianakis wrote:
+static bool
On Tue, Jun 27, 2017 at 01:45:40PM +0100, Stefan Hajnoczi wrote:
On Mon, Jun 26, 2017 at 07:26:41PM +0300, Manos Pitsidianakis wrote:
On Mon, Jun 26, 2017 at 03:30:55PM +0100, Stefan Hajnoczi wrote:
> > +bs->file = bdrv_open_child(NULL, optio
On Tue, Jun 27, 2017 at 02:08:54PM +0200, Alberto Garcia wrote:
On Fri 23 Jun 2017 02:46:53 PM CEST, Manos Pitsidianakis wrote:
This commit gathers ThrottleGroup membership details from
BlockBackendPublic into ThrottleGroupMember and refactors existing code
to use the structure.
Signed-off
On Mon, Jun 26, 2017 at 04:44:44PM +0100, Stefan Hajnoczi wrote:
On Fri, Jun 23, 2017 at 03:46:59PM +0300, Manos Pitsidianakis wrote:
-void blk_io_limits_disable(BlockBackend *blk)
+void blk_io_limits_disable(BlockBackend *blk, Error **errp)
{
-assert(blk
On Mon, Jun 26, 2017 at 03:52:34PM +0100, Stefan Hajnoczi wrote:
On Fri, Jun 23, 2017 at 03:46:56PM +0300, Manos Pitsidianakis wrote:
+static bool throttle_group_exists(const char *name)
+{
+ThrottleGroup *iter;
+bool ret = false;
+
+qemu_mutex_lock(_groups_lock);
Not sure
On Mon, Jun 26, 2017 at 03:30:55PM +0100, Stefan Hajnoczi wrote:
+bs->file = bdrv_open_child(NULL, options, "file",
+bs, _file, false, _err);
+
+if (local_err) {
+error_propagate(errp, local_err);
+return -EINVAL;
+}
+
+
On Mon, Jun 26, 2017 at 03:30:55PM +0100, Stefan Hajnoczi wrote:
+static BlockDriver bdrv_throttle = {
+.format_name= "throttle",
+.protocol_name = "throttle",
+.instance_size = sizeof(ThrottleGroupMember),
+
+
On Mon, Jun 26, 2017 at 03:52:34PM +0100, Stefan Hajnoczi wrote:
Have you seen iothread.c:qmp_query_iothreads()? All objects are put
into a container (the parent object), so you can just iterate over its
children. There's no need for a separate list because QOM already has
all the objects.
On Mon, Jun 26, 2017 at 02:36:11PM +0100, Stefan Hajnoczi wrote:
On Fri, Jun 23, 2017 at 03:46:54PM +0300, Manos Pitsidianakis wrote:
timer_cb() needs to know about the current Aio context of the throttle
request that is woken up. In order to make ThrottleGroupMember backend
agnostic
On Fri, Jun 23, 2017 at 03:46:55PM +0300, Manos Pitsidianakis wrote:
+static QemuOptsList throttle_opts = {
+.name = "throttle",
+.head = QTAILQ_HEAD_INITIALIZER(throttle_opts.head),
+.desc = {
+{
+.name = QEMU_OPT_IOPS_TOTAL,
throttle_opts
This commit gathers ThrottleGroup membership details from
BlockBackendPublic into ThrottleGroupMember and refactors existing code
to use the structure.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/block-backend.c | 66 +
block/
=...
The configuration flags and semantics are identical to the hardcoded
throttling ones.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/Makefile.objs | 1 +
block/throttle.c| 427
include/qemu/th
of the
device. The legacy throttle node is managed by the legacy interface
completely. More advanced configurations with the filter drive are
possible using the QMP API, but these will be ignored by the legacy
interface.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
tests/qemu-iotests/184 | 144 +
tests/qemu-iotests/184.out | 31 ++
tests/qemu-iotests/group | 1 +
3 files changed, 176 insertions(+)
create mode 100755 tests/qemu-i
Allow passing a QDict *options parameter to bdrv_new_open_driver() so
that it can be used if a driver needs it upon creation. The previous
behaviour (empty bs->options and bs->explicit_options) remains when
options is NULL.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
object
- convert legacy interface to use the driver
- add tests
Manos Pitsidianakis (8):
block: move ThrottleGroup membership to ThrottleGroupMember
block: Add aio_context field in ThrottleGroupMember
block: add throttle block filter driver
block: convert ThrottleGroup to object with QOM
This is needed to configure throttle filter driver nodes with QAPI.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
qapi/block-core.json | 19 ++-
1 file changed, 18 insertions(+), 1 deletion(-)
diff --git a/qapi/block-core.json b/qapi/block-core.json
timer_cb() needs to know about the current Aio context of the throttle
request that is woken up. In order to make ThrottleGroupMember backend
agnostic, this information is stored in an aio_context field instead of
accessing it from BlockBackend.
Signed-off-by: Manos Pitsidianakis <e
ThrottleGroup is converted to an object to allow easy runtime
configuration of throttling filter nodes in the BDS graph using QOM.
Signed-off-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
---
block/throttle-groups.c | 351
include/qemu/thro
On Sun, Jun 18, 2017 at 10:28:13AM +0200, Tobias Schramm wrote:
@@ -1469,6 +1472,28 @@ static int local_parse_opts(QemuOpts *opts, struct
FsDriverEntry *fse)
fse->path = g_strdup(path);
+fse->fmask = SM_LOCAL_MODE_BITS;
+if (fmask) {
+mask = strtol(fmask, NULL, 0);
(Use
Reviewed-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
On Fri, Jun 16, 2017 at 07:37:15PM +0200, Kevin Wolf wrote:
All functions that are marked coroutine_fn can directly call the
bdrv_co_* version of functions instead of going through the wrapper.
Signed-off-by: Kevin Wolf <kw...@r
Reviewed-by: Manos Pitsidianakis <el13...@mail.ntua.gr>
On Fri, Jun 16, 2017 at 07:37:10PM +0200, Kevin Wolf wrote:
Most of the qed code is now synchronous and matches the coroutine model.
One notable exception is the serialisation between requests which can
still schedule a callback. Bef
501 - 600 of 622 matches
Mail list logo