From: Klaus Jensen
There is no need to extract the format command parameters for each
namespace. Move it to the entry point.
Reviewed-by: Keith Busch
Signed-off-by: Klaus Jensen
---
hw/nvme/ctrl.c | 31 ++-
1 file changed, 18 insertions(+), 13 deletions(-)
diff
From: Naveen Nagar
This adds support for one possible new protection information format
introduced in TP4068 (and integrated in NVMe 2.0): the 64-bit CRC guard
and 48-bit reference tag. This version does not support storage tags.
Like the CRC16 support already present, this uses a software
From: Naveen Nagar
Add support for up to 64 LBA formats through the LBAFEE field of the
Host Behavior Support feature.
Reviewed-by: Keith Busch
Signed-off-by: Naveen Nagar
Signed-off-by: Klaus Jensen
---
hw/nvme/ctrl.c | 34 +++---
hw/nvme/ns.c | 15
From: Klaus Jensen
A subsequent patch will introduce a new tuple size; so add a helper and
use that instead of sizeof() and magic numbers.
Reviewed-by: Keith Busch
Signed-off-by: Klaus Jensen
---
hw/nvme/ctrl.c | 14 --
hw/nvme/dif.c | 16
hw/nvme/dif.h | 5
From: Naveen Nagar
Add support for getting and setting the Host Behavior Support feature.
Reviewed-by: Keith Busch
Signed-off-by: Naveen Nagar
Signed-off-by: Klaus Jensen
---
hw/nvme/ctrl.c | 8
hw/nvme/nvme.h | 4 +++-
include/block/nvme.h | 9 +
3 files
From: Klaus Jensen
Move dif/pi data structures and inlines to dif.h.
Reviewed-by: Keith Busch
Signed-off-by: Klaus Jensen
---
hw/nvme/ctrl.c | 1 +
hw/nvme/dif.c | 1 +
hw/nvme/dif.h | 53 ++
hw/nvme/nvme.h | 50
From: Klaus Jensen
Hi Peter,
Last round of hw/nvme updates for v7.0.
The following changes since commit 64ada298b98a51eb2512607f6e6180cb330c47b1:
Merge remote-tracking branch 'remotes/legoater/tags/pull-ppc-20220302' into
staging (2022-03-02 12:38:46 +)
are available in the Git
On Thu, Mar 03, 2022 at 03:59:02PM -0500, John Snow wrote:
> Fority compare_images() to be more discerning about the status codes it
Fortify
> receives. If qemu_img() returns an exit code that implies it didn't
> actually perform the comparison, treat that as an exceptional
> circumstance and
On Thu, Mar 03, 2022 at 03:59:01PM -0500, John Snow wrote:
> re-write qemu_img() as a function that will by default raise a
> VerboseProcessException (extended from CalledProcessException) on
> non-zero return codes. This will produce a stack trace that will show
> the command line arguments and
On Thu, Mar 03, 2022 at 03:58:59PM -0500, John Snow wrote:
> This adds an Exception that extends the Python stdlib
> subprocess.CalledProcessError.
>
> The difference is that the str() method of this exception also adds the
> stdout/stderr logs. In effect, if this exception goes unhandled, Python
On 3/3/22 06:55, Peter Maydell wrote:
Alternately, force size == 1, so that we always get a non-NULL value that can
be freed.
That's a change on the POSIX side as well, of course.
Yes, I had a look at what actual malloc() implementations tend
to do, and the answer seems to be that forcing
On Thu, Mar 03, 2022 at 04:03:22PM +, Daniel P. Berrangé wrote:
> When using the --list option, qemu-nbd acts as an NBD client rather
> than a server. As such when using TLS, it has a need to validate
> the server certificate. This adds a --tls-hostname option which can
> be used to override
On Thu, Mar 03, 2022 at 03:58:58PM -0500, John Snow wrote:
> >>> print(add_visual_margin(msg, width=72, name="Commit Message"))
> ┏━ Commit Message ━━
> ┃ add_visual_margin() takes a chunk of text and wraps it in a visual
> ┃ container that
On Thu, Mar 03, 2022 at 05:48:14PM +0100, Hanna Reitz wrote:
> 185 tests quitting qemu while a block job is active. It does not
> specifically test quitting qemu while a mirror or active commit job is
> in its READY phase.
>
> Add two test cases for this, where we respectively mirror or commit
On Thu, Mar 03, 2022 at 05:48:13PM +0100, Hanna Reitz wrote:
> To implement this, we reuse the existing daemonizing functions from the
> system emulator, which mainly do the following:
> - Fork off a child process, and set up a pipe between parent and child
> - The parent process waits until the
On Thu, Mar 03, 2022 at 05:48:12PM +0100, Hanna Reitz wrote:
> In contrast to qemu-nbd (where it is called --fork) and the system
> emulator, QSD does not have a --daemonize switch yet. Just like them,
> QSD allows setting up block devices and exports on the command line.
> When doing so, it is
On Thu, Mar 03, 2022 at 04:03:21PM +, Daniel P. Berrangé wrote:
> When connecting to an NBD server with TLS and x509 credentials,
> the client must validate the hostname it uses for the connection,
> against that published in the server's certificate. If the client
> is tunnelling its
Fority compare_images() to be more discerning about the status codes it
receives. If qemu_img() returns an exit code that implies it didn't
actually perform the comparison, treat that as an exceptional
circumstance and force the caller to be aware of the peril.
If a negative test is desired
qemu_img() returning zero ought to be the rule, not the
exception. Remove all explicit checks against the condition in
preparation for making non-zero returns an Exception.
Signed-off-by: John Snow
Reviewed-by: Eric Blake
---
tests/qemu-iotests/163 | 9 +++--
This adds an Exception that extends the Python stdlib
subprocess.CalledProcessError.
The difference is that the str() method of this exception also adds the
stdout/stderr logs. In effect, if this exception goes unhandled, Python
will print the output in a visually distinct wrapper to the terminal
>>> print(add_visual_margin(msg, width=72, name="Commit Message"))
┏━ Commit Message ━━
┃ add_visual_margin() takes a chunk of text and wraps it in a visual
┃ container that force-wraps to a specified width. An optional title
┃ label may be
re-write qemu_img() as a function that will by default raise a
VerboseProcessException (extended from CalledProcessException) on
non-zero return codes. This will produce a stack trace that will show
the command line arguments and return code from the failed process run.
Users that want something
This is kinda-sorta V3-ish of a series I started in response to Thomas
Huth's encountering a failure in qemu-img because of missing zstd
support. This series changes the qemu_img() function in iotests.py to
one that raises an Exception on non-zero return code by default.
Alongside this, the
On Thu, Mar 03, 2022 at 05:48:11PM +0100, Hanna Reitz wrote:
> The daemonizing functions in os-posix (os_daemonize() and
> os_setup_post()) only daemonize the process if the static `daemonize`
> variable is set. Right now, it can only be set by os_parse_cmd_args().
>
> In order to use
On Thu, Mar 03, 2022 at 04:03:20PM +, Daniel P. Berrangé wrote:
> In
>
> commit a71d597b989fd701b923f09b3c20ac4fcaa55e81
> Author: Vladimir Sementsov-Ogievskiy
> Date: Thu Jun 10 13:08:00 2021 +0300
>
> block/nbd: reuse nbd_co_do_establish_connection() in nbd_open()
>
> the use
On Thu, Mar 03, 2022 at 04:03:19PM +, Daniel P. Berrangé wrote:
> Currently the TLS session object assumes that the caller will always
> provide a hostname when using x509 creds on a client endpoint. This
> relies on the caller to detect and report an error if the user has
> configured QEMU
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
tests/qemu-iotests/tests/image-fleecing | 125 +++-
tests/qemu-iotests/tests/image-fleecing.out | 63 ++
2 files changed, 156 insertions(+), 32 deletions(-)
diff --git a/tests/qemu-iotests/tests/image-fleecing
The new block driver simply utilizes snapshot-access API of underlying
block node.
In further patches we want to use it like this:
[guest] [NBD export]
||
| root | root
v file v
Split intersecting-requests functionality out of block-copy to be
reused in copy-before-write filter.
Note: while being here, fix tiny typo in MAINTAINERS.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
include/block/reqlist.h | 67 +++
Note that reads zero areas (not dirty in the bitmap) fails, that's
correct.
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
tests/qemu-iotests/tests/image-fleecing | 38 +++---
tests/qemu-iotests/tests/image-fleecing.out | 84 +
2 files changed, 113 insertions(+), 9
Add new block driver handlers and corresponding generic wrappers.
It will be used to allow copy-before-write filter to provide
reach fleecing interface in further commit.
In future this approach may be used to allow reading qcow2 internal
snapshots, for example to export them through NBD.
Split block_copy_reset() out of block_copy_reset_unallocated() to be
used separately later.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
include/block/block-copy.h | 1 +
block/block-copy.c | 21 +
2 files changed, 14 insertions(+), 8
Let's reuse convenient helper.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
block/reqlist.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/block/reqlist.c b/block/reqlist.c
index 5e320ba649..09fecbd48c 100644
--- a/block/reqlist.c
+++
Add helper that returns both status and output, to be used in the
following commit
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
tests/qemu-iotests/iotests.py | 3 +++
1 file changed, 3 insertions(+)
diff --git a/tests/qemu-iotests/iotests.py b/tests/qemu-iotests/iotests.py
index
Add a convenient function similar with bdrv_block_status() to get
status of dirty bitmap.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
include/block/dirty-bitmap.h | 2 ++
include/qemu/hbitmap.h | 12
block/dirty-bitmap.c | 6 ++
Current scheme of image fleecing looks like this:
[guest][NBD export]
| |
|root | root
v v
[copy-before-write] -> [temp.qcow2]
| target |
|file
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
tests/qemu-iotests/tests/image-fleecing | 64 -
tests/qemu-iotests/tests/image-fleecing.out | 76 -
2 files changed, 120 insertions(+), 20 deletions(-)
diff --git
This brings "incremental" mode to copy-before-write filter: user can
specify bitmap so that filter will copy only "dirty" areas.
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
qapi/block-core.json | 10 +++-
block/copy-before-write.c | 51 ++-
2
This will be used in the following commit to bring "incremental" mode
to copy-before-write filter.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
include/block/block-copy.h | 1 +
block/block-copy.c | 14 +-
block/copy-before-write.c | 2 +-
3
That simplifies handling failure in existing code and in further new
usage of bdrv_merge_dirty_bitmap().
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Hanna Reitz
---
include/block/dirty-bitmap.h| 2 +-
block/dirty-bitmap.c| 9 +++--
Add function to wait for all intersecting requests.
To be used in the further commit.
Signed-off-by: Vladimir Sementsov-Ogievskiy
Reviewed-by: Nikita Lapshin
Reviewed-by: Hanna Reitz
---
include/block/reqlist.h | 8
block/reqlist.c | 8
2 files changed, 16
We are going to complicate bitmap initialization in the further
commit. And in future, backup job will be able to work without filter
(when source is immutable), so we'll need same bitmap initialization in
copy-before-write filter and in backup job. So, it's reasonable to do
it in block-copy.
v6:
11: add comment
15: limit to qcow2 with unsupported compat
fix style
16: fix style
change log('Backup finished ...') to assertion and comment
Vladimir Sementsov-Ogievskiy (16):
block/block-copy: move copy_bitmap initialization to
block_copy_state_new()
block/dirty-bitmap:
03.03.2022 13:58, Hanna Reitz wrote:
On 28.02.22 12:39, Vladimir Sementsov-Ogievskiy wrote:
Signed-off-by: Vladimir Sementsov-Ogievskiy
---
tests/qemu-iotests/tests/image-fleecing | 120 ++--
tests/qemu-iotests/tests/image-fleecing.out | 63 ++
2 files changed,
03.03.2022 14:11, Hanna Reitz wrote:
On 03.03.22 12:05, Hanna Reitz wrote:
On 28.02.22 12:39, Vladimir Sementsov-Ogievskiy wrote:
The new block driver simply utilizes snapshot-access API of underlying
block node.
In further patches we want to use it like this:
[guest] [NBD
Am 16.02.2022 um 11:53 hat Hanna Reitz geschrieben:
> bdrv_refresh_limits() recurses down to the node's children. That does
> not seem necessary: When we refresh limits on some node, and then
> recurse down and were to change one of its children's BlockLimits, then
> that would mean we noticed
On Sun, 27 Feb 2022 at 18:36, Richard Henderson
wrote:
>
> On 2/27/22 02:54, Peter Maydell wrote:
> >>> +if (size) {
> >>> +ptr = _aligned_malloc(size, alignment);
> >>> +} else {
> >>> +ptr = NULL;
> >>> +}
> >>
> >> Oh, should we set errno to something here?
> >>
185 tests quitting qemu while a block job is active. It does not
specifically test quitting qemu while a mirror or active commit job is
in its READY phase.
Add two test cases for this, where we respectively mirror or commit to
an external QSD instance, which provides a throttled block device.
To implement this, we reuse the existing daemonizing functions from the
system emulator, which mainly do the following:
- Fork off a child process, and set up a pipe between parent and child
- The parent process waits until the child sends a status byte over the
pipe (0 means that the child was
In contrast to qemu-nbd (where it is called --fork) and the system
emulator, QSD does not have a --daemonize switch yet. Just like them,
QSD allows setting up block devices and exports on the command line.
When doing so, it is often necessary for whoever invoked the QSD to wait
until these
The daemonizing functions in os-posix (os_daemonize() and
os_setup_post()) only daemonize the process if the static `daemonize`
variable is set. Right now, it can only be set by os_parse_cmd_args().
In order to use os_daemonize() and os_setup_post() from the storage
daemon to have it be
Hi,
v1 cover letter:
https://lists.nongnu.org/archive/html/qemu-block/2021-12/msg00499.html
In v2, I followed Vladimir’s suggestion to look into whether we could
reuse os_daemonize(). Indeed we can, and it makes patch 3 (formerly 2)
much simpler!
I decided to leave patch 2 (formerly 1)
Using standard filters is more future proof than rolling our own.
Signed-off-by: Daniel P. Berrangé
---
tests/qemu-iotests/233 | 29 -
tests/qemu-iotests/233.out | 9 -
2 files changed, 16 insertions(+), 22 deletions(-)
diff --git
This validates that connections to an NBD server running on a UNIX
socket can use TLS with pre-shared keys (PSK).
Signed-off-by: Daniel P. Berrangé
---
tests/qemu-iotests/233| 28
tests/qemu-iotests/233.out| 17 +
The TLS usage for NBD was restricted to IP sockets because validating
x509 certificates requires knowledge of the hostname that the client
is connecting to.
TLS does not have to use x509 certificates though, as PSK (pre-shared
keys) provide an alternative credential option. These have no
When using the --list option, qemu-nbd acts as an NBD client rather
than a server. As such when using TLS, it has a need to validate
the server certificate. This adds a --tls-hostname option which can
be used to override the default hostname used for certificate
validation.
Signed-off-by: Daniel
When connecting to an NBD server with TLS and x509 credentials,
the client must validate the hostname it uses for the connection,
against that published in the server's certificate. If the client
is tunnelling its connection over some other channel, however, the
hostname it uses may not match the
In
commit a71d597b989fd701b923f09b3c20ac4fcaa55e81
Author: Vladimir Sementsov-Ogievskiy
Date: Thu Jun 10 13:08:00 2021 +0300
block/nbd: reuse nbd_co_do_establish_connection() in nbd_open()
the use of the 'hostname' field from the BDRVNBDState struct was
lost, and 'nbd_connect' just
Introduce a filter for the output of qemu-nbd export list so it can be
reused in multiple tests.
The filter is a bit more permissive that what test 241 currently uses,
as its allows printing of the export count, along with any possible
error messages that might be emitted.
Signed-off-by: Daniel
This validates that connections to an NBD server running on a UNIX
socket can use TLS, and require a TLS hostname override to pass
certificate validation.
Signed-off-by: Daniel P. Berrangé
---
tests/qemu-iotests/233 | 24
tests/qemu-iotests/233.out | 15
This validates that connections to an NBD server where the certificate
hostname does not match will fail. It further validates that using the
new 'tls-hostname' override option can solve the failure.
Signed-off-by: Daniel P. Berrangé
---
tests/qemu-iotests/233| 18 ++
Some tests will want to use 'localhost' instead of '127.0.0.1', and
some will use the image options syntax rather than the classic URI
syntax.
Signed-off-by: Daniel P. Berrangé
---
tests/qemu-iotests/common.filter | 4
1 file changed, 4 insertions(+)
diff --git
When developing an I/O test it is typical to add some logic to the
test script, run it to view the output diff, and then apply the
output diff to the reference file. This can be drastically simplified
by letting the test runner update the reference file in place.
By setting
Currently the TLS session object assumes that the caller will always
provide a hostname when using x509 creds on a client endpoint. This
relies on the caller to detect and report an error if the user has
configured QEMU with x509 credentials on a UNIX socket. The migration
code has such a check,
This series was principally motivated by a desire to enabl use of TLS
on non-TCP transports. For x509 certificates this means we need a way
to set the hostname to use for validation. This also lets us override
the hostname when connecting on a TCP transport that is tunnelled or
port-forwarded. It
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/block.c b/block.c
index 5afdbe3060..01811d6519 100644
--- a/block.c
+++ b/block.c
@@ -1497,7 +1497,7 @@ const BdrvChildClass child_of_bds = {
AioContext
Signed-off-by: Emanuele Giuseppe Esposito
---
include/block/block_int-common.h | 81 ++--
1 file changed, 47 insertions(+), 34 deletions(-)
diff --git a/include/block/block_int-common.h b/include/block/block_int-common.h
index f05ebb0da3..5a04c778e4 100644
---
Signed-off-by: Emanuele Giuseppe Esposito
---
job.c | 10 ++
1 file changed, 10 insertions(+)
diff --git a/job.c b/job.c
index 54db80df66..075c6f3a20 100644
--- a/job.c
+++ b/job.c
@@ -381,6 +381,8 @@ void job_ref(Job *job)
void job_unref(Job *job)
{
+GLOBAL_STATE_CODE();
+
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c| 17 +
block/create.c | 2 ++
2 files changed, 19 insertions(+)
diff --git a/block.c b/block.c
index 4a3447b2a0..5afdbe3060 100644
--- a/block.c
+++ b/block.c
@@ -529,6 +529,7 @@ static void coroutine_fn
Assertions in the callers of the function pointrs are already
added by previous patches.
Signed-off-by: Emanuele Giuseppe Esposito
Reviewed-by: Stefan Hajnoczi
Reviewed-by: Philippe Mathieu-Daudé
---
include/sysemu/block-backend-common.h | 28 ++-
1 file changed, 23
Similar to the header split, also the function pointers in BlockDriver
can be split in I/O and global state.
Signed-off-by: Emanuele Giuseppe Esposito
---
include/block/block_int-common.h | 445 ---
1 file changed, 237 insertions(+), 208 deletions(-)
diff --git
copy-before-write functions always run under BQL.
Signed-off-by: Emanuele Giuseppe Esposito
---
block/copy-before-write.c | 2 ++
block/copy-before-write.h | 7 +++
2 files changed, 9 insertions(+)
diff --git a/block/copy-before-write.c b/block/copy-before-write.c
index
Signed-off-by: Emanuele Giuseppe Esposito
---
blockjob.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/blockjob.c b/blockjob.c
index d79a52d204..4868453d74 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -62,6 +62,7 @@ static bool is_block_job(Job *job)
BlockJob
blockdev functions run always under the BQL lock.
Signed-off-by: Emanuele Giuseppe Esposito
---
include/sysemu/blockdev.h | 13 ++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/include/sysemu/blockdev.h b/include/sysemu/blockdev.h
index f9fb54d437..3211b16513 100644
blockjob functions run always under the BQL lock.
Signed-off-by: Emanuele Giuseppe Esposito
---
include/block/blockjob.h | 29 ++---
1 file changed, 22 insertions(+), 7 deletions(-)
diff --git a/include/block/blockjob.h b/include/block/blockjob.h
index
Now that we "covered" the three main cases where the
permission API was being used under BQL (fuse,
amend and invalidate_cache), we can safely assert for
the permission functions implemented in block.c
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c | 12
1 file changed, 12
Snapshots run also under the BQL, so they all are
in the global state API. The aiocontext lock that they hold
is currently an overkill and in future could be removed.
Signed-off-by: Emanuele Giuseppe Esposito
---
block/snapshot.c | 28
The job API will be handled separately in another serie.
Signed-off-by: Emanuele Giuseppe Esposito
---
include/qemu/job.h | 22 ++
1 file changed, 22 insertions(+)
diff --git a/include/qemu/job.h b/include/qemu/job.h
index 6e67b6977f..c105b31076 100644
---
Signed-off-by: Emanuele Giuseppe Esposito
---
blockjob.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/blockjob.c b/blockjob.c
index 10815a89fe..d79a52d204 100644
--- a/blockjob.c
+++ b/blockjob.c
@@ -84,6 +84,7 @@ BlockJob *block_job_get(const char *id)
void block_job_free(Job *job)
Following the assertion derived from the API split,
propagate the assertion also in the static functions.
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c | 46 ++-
block/block-backend.c | 3 +++
2 files changed, 48 insertions(+), 1
We want to be sure that the functions that write the child and
parent list of a bs are under BQL and drain.
BQL prevents from concurrent writings from the GS API, while
drains protect from I/O.
TODO: drains are missing in some functions using this assert.
Therefore a proper assertion will fail.
block coroutines functions run in different aiocontext, and are
not protected by the BQL. Therefore are I/O.
On the other side, generated_co_wrapper functions use BDRV_POLL_WHILE,
meaning the caller can either be the main loop or a specific iothread.
Signed-off-by: Emanuele Giuseppe Esposito
Signed-off-by: Emanuele Giuseppe Esposito
---
block/block-backend.c | 3 +++
blockdev.c| 16
2 files changed, 19 insertions(+)
diff --git a/block/block-backend.c b/block/block-backend.c
index 2ab1274dfe..bf77c4a8fa 100644
--- a/block/block-backend.c
+++
Since the I/O functions are not many, keep a single file.
Also split the function pointers in BlockJobDriver.
Signed-off-by: Emanuele Giuseppe Esposito
Reviewed-by: Stefan Hajnoczi
---
include/block/blockjob_int.h | 28
1 file changed, 28 insertions(+)
diff --git
All the global state (GS) API functions will check that
qemu_in_main_thread() returns true. If not, it means
that the safety of BQL cannot be guaranteed, and
they need to be moved to I/O.
Signed-off-by: Emanuele Giuseppe Esposito
---
block/block-backend.c | 78
When invoked from the main loop, this function is the same
as qemu_mutex_iothread_locked, and returns true if the BQL is held.
When invoked from iothreads or tests, it returns true only
if the current AioContext is the Main Loop.
This essentially just extends qemu_mutex_iothread_locked to work
Similarly to the previous patch, split block_int.h
in block_int-io.h and block_int-global-state.h
block_int-common.h contains the structures shared between
the two headers, and the functions that can't be categorized as
I/O or global state.
Assertions are added in the next patch.
Signed-off-by:
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c | 15 +++
block/backup.c | 1 +
block/block-backend.c | 3 +++
block/commit.c | 2 ++
block/dirty-bitmap.c| 1 +
block/io.c |
Mark all I/O functions with IO_CODE, and all "I/O OR GS" with
IO_OR_GS_CODE.
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c | 14 +-
block/block-backend.c| 2 ++
block/dirty-bitmap.c | 3 +++
block/io.c | 13
block.h currently contains a mix of functions:
some of them run under the BQL and modify the block layer graph,
others are instead thread-safe and perform I/O in iothreads.
Some others can only be called by either the main loop or the
iothread running the AioContext (and not other iothreads),
and
Mark all I/O functions with IO_CODE, and all "I/O OR GS" with
IO_OR_GS_CODE.
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c | 38 ++-
block/dirty-bitmap.c | 1 +
block/io.c | 43 ++--
Mark all I/O functions with IO_CODE, and all "I/O OR GS" with
IO_OR_GS_CODE.
Signed-off-by: Emanuele Giuseppe Esposito
---
block/block-backend.c | 58 +++
include/sysemu/block-backend-io.h | 2 ++
2 files changed, 60 insertions(+)
diff --git
Similarly to the previous patches, split block-backend.h
in block-backend-io.h and block-backend-global-state.h
In addition, remove "block/block.h" include as it seems
it is not necessary anymore, together with "qemu/iov.h"
block-backend-common.h contains the structures shared between
the two
Currently, block layer APIs like block.h contain a mix of
functions that are either running in the main loop and under the
BQL, or are thread-safe functions and run in iothreads performing I/O.
The functions running under BQL also take care of modifying the
block graph, by using drain and/or
Allow writable exports to get BLK_PERM_RESIZE permission
from creation, in fuse_export_create().
In this way, there is no need to give the permission in
fuse_do_truncate(), which might be run in an iothread.
Permissions should be set only in the main thread, so
in any case if an iothread tries to
All the global state (GS) API functions will check that
qemu_in_main_thread() returns true. If not, it means
that the safety of BQL cannot be guaranteed, and
they need to be moved to I/O.
Signed-off-by: Emanuele Giuseppe Esposito
---
block.c| 124
Righ now, IO_CODE and IO_OR_GS_CODE are nop, as there isn't
really a way to check that a function is only called in I/O.
On the other side, we can use qemu_in_main_thread() to check if
we are in the main loop.
The usage of macros makes easy to extend them in the future without
making changes in
The thread pool regulates itself: when idle, it kills threads until
empty, when in demand, it creates new threads until full. This behaviour
doesn't play well with latency sensitive workloads where the price of
creating a new thread is too high. For example, when paired with qemu's
'-mlock', or
Introduce the 'event-loop-base' abstract class, it'll hold the
properties common to all event loops and provide the necessary hooks for
their creation and maintenance. Then have iothread inherit from it.
EventLoopBaseClass is defined as user creatable and provides a hook for
its children to
Upon freeing a thread pool we need to get rid of any remaining worker.
This is achieved by setting the thread pool's topping flag, waking the
workers up, and waiting for them to exit one by one. The problem is that
currently all this process happens with the thread pool lock held,
effectively
1 - 100 of 121 matches
Mail list logo
