On 11/01/2018 12:29, Laurent Vivier wrote:
> On 06/01/2018 01:47, Jose Ricardo Ziviani wrote:
>> Power9 supports 4 HW threads/core but it's possible to emulate
>> doorbells to implement virtual SMT. KVM has the KVM_CAP_PPC_SMT_POSSIBLE
>> which returns a bitmap with all SMT modes supported by the h
On Wed, 10 Jan 2018 00:39:02 +0800
Jay Zhou wrote:
Jay,
considering how non trivial touched code is, would you mind
first adding 'make check' testcases for success/failure paths
that you are touching?
It would help with merging and ensure that future changes to
vhost won't break memslots handling
On Thu, Jan 11, 2018 at 01:46:38PM +0100, Max Reitz wrote:
> On 2018-01-08 14:52, Eric Blake wrote:
> > On 01/07/2018 06:23 AM, Richard Palethorpe wrote:
> >> Add QAPI wrapper functions for the existing snapshot functionality. These
> >> functions behave the same way as the HMP savevm, loadvm and d
On Thu, Jan 11, 2018 at 11:25:08AM +0100, Igor Mammedov wrote:
> On Wed, 10 Jan 2018 19:48:00 -0200
> Eduardo Habkost wrote:
>
> > On Wed, Jan 10, 2018 at 01:30:29PM -0800, Alistair Francis wrote:
> > > On Thu, Dec 28, 2017 at 6:59 AM, Eduardo Habkost
> > > wrote:
> > > > On Thu, Dec 28, 2017
On 11/01/2018 09:24, Pavel Dovgalyuk wrote:
> Accidentally sent draft version of patches to the mailing list.
> Please consider this one as the correct one.
>
> This set of patches includex fixes from Alex Bennée for fixing
> BQL and replay locks after inventing the MTTCG. It also includes some
>
On 11/01/2018 09:27, Pavel Dovgalyuk wrote:
> +} else {
> +qemu_notify_event();
> +}
Before this patch, what would do the qemu_notify_event?
Thanks,
Paolo
> /* Does not need atomic_mb_set because a spurious wakeup is okay. */
Except for round-robin TCG, every other accelerator is using more or
less the same code around qemu_wait_io_event_common. The exception
is HAX, which also has to eat the dummy APC that is queued by
qemu_cpu_kick_thread.
We can add the SleepEx call to qemu_wait_io_event under "if
(!tcg_enabled())"
On 11/01/2018 09:26, Pavel Dovgalyuk wrote:
> From: Alex Bennée
>
> We only really need to grab the lock for initial setup (so we don't
> race with the thread-spawning thread). After that we can drop the lock
> for the whole main loop and only grab it for waiting for IO events.
>
> There is a sl
* Juan Quintela (quint...@redhat.com) wrote:
> We move to use migration_set_parameter() for everything.
>
> Signed-off-by: Juan Quintela
Reviewed-by: Dr. David Alan Gilbert
> ---
> tests/migration-test.c | 33 -
> 1 file changed, 8 insertions(+), 25 deletions(-
On 2018-01-08 14:52, Eric Blake wrote:
> On 01/07/2018 06:23 AM, Richard Palethorpe wrote:
>> Add QAPI wrapper functions for the existing snapshot functionality. These
>> functions behave the same way as the HMP savevm, loadvm and delvm
>> commands. This will allow applications, such as OpenQA, to
On 11/01/2018 09:27, Pavel Dovgalyuk wrote:
> Virtual clock is wapred from iothread and vcpu thread. When the hardware
> events associated with warp checkpoint, then interrupt delivering may be
> non-deterministic if checkpoint is processed in different threads in record
> and replay.
> This patch
Thanks Fabien,
Now, as a side question, could you tell me which reference LEON3
platform is implemented by Qemu in leon3_generic?
It doesn't seem to match the one emulated by tsim.
Thanks.
JC
Le 2018-01-11 12:48, Fabien Chouteau a écrit :
On 10/01/2018 21:43, Jean-Christophe Dubois wrote:
> -Original Message-
> From: Jason Wang [mailto:jasow...@redhat.com]
> Sent: Thursday, January 11, 2018 6:30 PM
> To: Zhoujian (jay) ; qemu-devel@nongnu.org
> Cc: Huangweidong (C) ; m...@redhat.com; wangxin (U)
> ; Gonglei (Arei) ;
> imamm...@redhat.com; Liuzhe (Ahriy, Euler)
> Subject: Re
On Fri, 5 Jan 2018 14:07:29 +1100
David Gibson wrote:
> On Thu, Jan 04, 2018 at 06:47:18PM +0100, Greg Kurz wrote:
> > On Thu, 4 Jan 2018 15:24:05 +1100
> > David Gibson wrote:
> >
> > > Currently the pseries machine sets the compatibility mode for the
> > > guest's cpus in two places: 1) at
I'm not sure, the current implementation is multi-threaded so I won't
hit this bug if its still present. If I can find the time I will make a
proof of concept and test if the bug is still present.
--
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribe
On 9 January 2018 at 18:50, Alistair Francis
wrote:
> Initial commit of the ZynqMP RTC device.
>
> Signed-off-by: Alistair Francis
Hi -- did you forget to send the cover letter for this series, or
did the mailing list server drop it? I only see the 3 patch mails...
thanks
-- PMM
On 3 January 2018 at 09:38, Juan Quintela wrote:
> Hi
>
> This are the changes for migration that are already reviewed.
>
> Please, apply.
>
> Later, Juan.
>
>
> The following changes since commit 281f327487c9c9b1599f93c589a408bbf4a651b8:
>
> Merge remote-tracking branch
> 'remotes/vivier/tags/
On 10/01/2018 21:43, Jean-Christophe Dubois wrote:
> With the LEON3 IRQ controller IRQs can be acknoledged 2 ways:
> * Explicitely by software writing to the CLEAR_OFFSET register
> * Implicitely when the procesor is done running the trap handler attached
> to the IRQ.
>
Thanks Jean-Christophe,
On 06/01/2018 01:47, Jose Ricardo Ziviani wrote:
> Power9 supports 4 HW threads/core but it's possible to emulate
> doorbells to implement virtual SMT. KVM has the KVM_CAP_PPC_SMT_POSSIBLE
> which returns a bitmap with all SMT modes supported by the host.
>
> Today, QEMU forces the SMT mode based
Wei Wang wrote:
> Michael, could we merge patch 3-5 first?
No! I'm repeatedly asking you to propose only VIRTIO_BALLOON_F_SG changes.
Please don't ignore me.
Patch 4 depends on patch 2. Thus, back to patch 2.
Your patch is trying to switch tell_host_sgs() and tell_host() based on
VIRTIO_BALLOO
On 9 January 2018 at 22:18, Mark Cave-Ayland
wrote:
> Thanks for the hint - while it didn't crash locally, I was certainly able to
> reproduce the above trace in valgrind.
>
> Turns out the issue was that thought I could move pci_register_bus() from
> realize to init in patch 10, but evidently not
On 2018年01月11日 00:14, Stefan Hajnoczi wrote:
Hi Wei,
I wanted to summarize the differences between the vhost-pci and
virtio-vhost-user approaches because previous discussions may have been
confusing.
vhost-pci defines a new virtio device type for each vhost device type
(net, scsi, blk). It th
* Juan Quintela (quint...@redhat.com) wrote:
> Also reorder code to not sleep when event already happened.
>
> Signed-off-by: Juan Quintela
> ---
> tests/migration-test.c | 29 ++---
> 1 file changed, 14 insertions(+), 15 deletions(-)
>
> diff --git a/tests/migration-tes
On Thu, 11 Jan 2018 04:04:18 +0100
Dong Jia Shi wrote:
> Hi Folks,
>
> Background
> ==
>
> Some days ago, we had a discussion on the topic of channel path
> virtualization.
> Ref:
> Subject: [PATCH 0/3] Channel Path realted CRW generation
> Message-Id: <20170727015418.85407-1-bjsdj...@
+-- On Mon, 25 Dec 2017, linzhecheng wrote --+
| --- a/hw/display/vga.c
| +++ b/hw/display/vga.c
| @@ -1279,6 +1279,10 @@ static void vga_draw_text(VGACommonState *s, int
full_update)
| cx_min = width;
| cx_max = -1;
| for(cx = 0; cx < width; cx++) {
| +if (s
Hi,
This series failed automatic build test. Please find the testing commands and
their output below. If you have docker installed, you can probably reproduce it
locally.
Type: series
Message-id: 20180111082452.27295.85707.stgit@pasha-VirtualBox
Subject: [Qemu-devel] [RFC PATCH v3 00/30] replay a
Hi,
This series seems to have some coding style problems. See output below for
more information:
Type: series
Message-id: 20180111082452.27295.85707.stgit@pasha-VirtualBox
Subject: [Qemu-devel] [RFC PATCH v3 00/30] replay additions
=== TEST SCRIPT BEGIN ===
#!/bin/bash
BASE=base
n=1
total=$(git
On 2018/1/11 17:53, Peter Maydell wrote:
>> we only support AArch64 kernel, not support AArch32
>> kernel or AArch32 user space.
> Nope, you must handle AArch32 EL1 correctly in some way, even if that
> is only "this guest CPU doesn't support RAS notification and we
> will not notify it". And you
* Juan Quintela (quint...@redhat.com) wrote:
> This way, it is like the rest of commands
>
> Signed-off-by: Juan Quintela
Reviewed-by: Dr. David Alan Gilbert
> ---
> tests/migration-test.c | 14 ++
> 1 file changed, 10 insertions(+), 4 deletions(-)
>
> diff --git a/tests/migratio
On 2018年01月11日 11:54, Zhoujian (jay) wrote:
Hi Jason,
-Original Message-
From: Jason Wang [mailto:jasow...@redhat.com]
Sent: Thursday, January 11, 2018 11:35 AM
To: Zhoujian (jay) ; qemu-devel@nongnu.org
Cc: Huangweidong (C) ; m...@redhat.com; wangxin (U)
; Gonglei (Arei) ;
imamm...@r
On Wed, 10 Jan 2018 19:48:00 -0200
Eduardo Habkost wrote:
> On Wed, Jan 10, 2018 at 01:30:29PM -0800, Alistair Francis wrote:
> > On Thu, Dec 28, 2017 at 6:59 AM, Eduardo Habkost
> > wrote:
> > > On Thu, Dec 28, 2017 at 02:39:31PM +0100, Igor Mammedov wrote:
> > >> On Fri, 22 Dec 2017 11:47
On 11 January 2018 at 10:15, Paolo Bonzini wrote:
> RAM contents are memcpy'd blindly during loadvm. I think that's what
> requires a tb_flush.
Ah, that makes sense. Could we have a comment documenting the
rationale, please? Something like:
/* vmload has just updated the content of RAM, bypassi
The loading time of a VM is quite significant when its virtio
devices uses a large amount of virt-queues (e.g. a virtio-serial
device with max_ports=511). Most of the time is spend in the
creation of all the required event notifiers (ioeventfd and memory
regions).
This patch pack all the changes t
On 10/01/2018 19:32, Peter Maydell wrote:
> On 10 January 2018 at 17:49, Richard Henderson
> wrote:
>> On 01/10/2018 05:48 AM, Pavel Dovgalyuk wrote:
>>> Flushing TB cache is required because TBs key in the cache may match
>>> different code which existed in the previous state.
>>>
>>> Signed-off-
On Thu, Jan 11, 2018 at 6:31 AM, Wei Wang wrote:
> On 01/11/2018 12:14 AM, Stefan Hajnoczi wrote:
>>
>> Hi Wei,
>> I wanted to summarize the differences between the vhost-pci and
>> virtio-vhost-user approaches because previous discussions may have been
>> confusing.
>>
>> vhost-pci defines a new
On 9 January 2018 at 18:17, Max Filippov wrote:
> Hi Peter,
>
> please pull the following batch of updates for the target/xtensa.
> Changes v2->v3:
> - Don't use g_malloc_n, use g_new instead.
> Changes v1->v2:
> - Drop no longer used function option_bits_enabled.
>
> The following changes since c
On 11 January 2018 at 05:59, gengdongjiu wrote:
> Hi Peter.
>
> On 2018/1/10 1:30, Peter Maydell wrote:
>> On 28 December 2017 at 05:54, Dongjiu Geng wrote:
>>> +
>>> +/* This exception comes from lower or current exception level. */
>>> +if ((cpsr & 0xf) == PSTATE_MODE_EL0t) {
>>
>> Thi
On Thu, Jan 11, 2018 at 08:06:13AM +, Mark Cave-Ayland wrote:
> On 11/01/18 04:08, Alexey Kardashevskiy wrote:
>
> > As stated in the 1ad9f0a464fe commit log, the returned entries are not
> > a while PTEG. It was not a problem before 1ad9f0a464fe as it would read
>
> s/while/whole/?
I alread
Hi,
> I wonder if it is reasonable for ui/input-linux.c to honour the same
> global '-ungrab' configuration option ?
No. Problem here is that input-linux is basically sniffing keyboard
input when the host owns the keyboard, so it can't prevent the hotkey
being seen by the host. Therefore I pi
On Wed, Jan 10, 2018 at 04:43:22PM +, Nir Soffer wrote:
> On Wed, Jan 10, 2018 at 4:04 PM Kashyap Chamarthy
> wrote:
[...]
> > Yes, for completness' sake, Nova upstream is already patched to use the
> > `qemu-img` '--force-share' flag that comes with QEMU >= 2.10.
> >
>
> What abut users ru
On Wed, 10 Jan 2018 18:21:57 -0800
Michael Clark wrote:
> The PLIC (Platform Level Interrupt Controller) device provides a
> parameterizable interrupt controller based on SiFive's PLIC specification.
>
> Signed-off-by: Michael Clark
> ---
> hw/riscv/sifive_plic.c | 554
> +
This patch adds a flag which denies recursive call of replay_checkpoint
function. Checkpoints may be accompanied by the hardware events. When event
is processed, virtual device may invoke timer modification functions that
also invoke the checkpoint function. This leads to infinite loop.
Signed-off
Virtual clock is wapred from iothread and vcpu thread. When the hardware
events associated with warp checkpoint, then interrupt delivering may be
non-deterministic if checkpoint is processed in different threads in record
and replay.
This patch disables event processing for clock warp checkpoint an
This patch adds error reporting when fwrite cannot completely
save the buffer to the file.
Signed-off-by: Pavel Dovgalyuk
--
v3: also check putc() return value
---
replay/replay-internal.c | 17 +++--
1 file changed, 15 insertions(+), 2 deletions(-)
diff --git a/replay/replay-in
From: Alex Bennée
This script allows analysis of mutex acquisition and hold times based
on a trace file. Given a trace control file of:
qemu_mutex_lock
qemu_mutex_locked
qemu_mutex_unlock
And running with:
$QEMU $QEMU_ARGS -trace events=./lock-trace
You can analyse the results with:
This patch skips cpu_exec when there is no CPU code to execute in replay mode.
Signed-off-by: Pavel Dovgalyuk
Signed-off-by: Maria Klimushenkova
---
cpus.c | 50 +--
include/sysemu/replay.h |3 +++
replay/replay.c | 13
From: Alex Bennée
The replay_mutex_lock/unlock/locked functions are now going to be used
for ensuring lock-step behaviour between the two threads. Make them
public API functions and also provide stubs for non-QEMU builds on
common paths.
Signed-off-by: Alex Bennée
Tested-by: Pavel Dovgalyuk
--
From: Alex Bennée
This introduces the qemu-gdb command "qemu timers" which will dump the
state of the main timers in the system.
Signed-off-by: Alex Bennée
Reviewed-by: Philippe Mathieu-Daudé
---
scripts/qemu-gdb.py |3 ++-
scripts/qemugdb/timers.py | 54 ++
From: Alex Bennée
Now instead of using the replay_lock to guard the output of the log we
now use it to protect the whole execution section. This replaces what
the BQL used to do when it was held during TCG execution.
We also introduce some rules for locking order - mainly that you
cannot take th
From: Alex Bennée
This script is a debugging tool for looking through the contents of a
replay log file. It is incomplete but should fail gracefully at events
it doesn't understand.
It currently understands two different log formats as the audio
record/replay support was merged during since MTTC
From: Alex Bennée
We only really need to grab the lock for initial setup (so we don't
race with the thread-spawning thread). After that we can drop the lock
for the whole main loop and only grab it for waiting for IO events.
There is a slight wrinkle for the round-robin TCG thread as we also
exp
From: Alex Bennée
Add a simple helper to dump lock state.
Signed-off-by: Alex Bennée
---
scripts/qemu-gdb.py|3 ++-
scripts/qemugdb/tcg.py | 46 ++
2 files changed, 48 insertions(+), 1 deletion(-)
create mode 100644 scripts/qemugdb/tcg.py
From: Alex Bennée
Now the only real need to hold the BQL is for when we sleep on the
cpu->halt conditional. The lock is actually dropped while the thread
sleeps so the actual window for contention is pretty small. This also
means we can remove the special case hack for exclusive work and
simply d
From: Alex Bennée
Signed-off-by: Alex Bennée
---
v1
- fix merge conflicts with existing tracing
- add trylock/cond_wait traces
---
include/qemu/thread.h| 19 +++
util/qemu-thread-posix.c | 21 -
util/trace-events|7 ---
3 files ch
From: Alex Bennée
This time commit 802f045a5f61b781df55e4492d896b4d20503ba7 broke the
replay file format. Also add a comment about this to
replay-internal.h.
Signed-off-by: Alex Bennée
Reviewed-off-by: Pavel Dovgalyuk
Acked-by: Paolo Bonzini
---
replay/replay-internal.h |2 +-
replay/rep
From: Pavel Dovgalyuk
This patch disables bdrv flush/drain in record/replay mode.
When block request is in the replay queue it cannot be processed
with drain/flush until it is found in the log.
Therefore vm should just stop leaving unfinished operations
in the queue.
Signed-off-by: Pavel Dovgaly
From: Alex Bennée
The powerctl code is run in the context of the vCPU changing power
state. It does not need the BQL to protect its changes.
Signed-off-by: Alex Bennée
---
target/arm/arm-powerctl.c |8
1 file changed, 8 deletions(-)
diff --git a/target/arm/arm-powerctl.c b/target
Replay mutex is held by vCPU thread and destroy function is called
from atexit of the main thread. Therefore we cannot destroy it safely.
Signed-off-by: Pavel Dovgalyuk
Acked-by: Paolo Bonzini
---
replay/replay.c |1 -
1 file changed, 1 deletion(-)
diff --git a/replay/replay.c b/replay/rep
This patch adds saving/restoring of the host clock field 'last'.
It is used in host clock calculation and therefore clock may
become incorrect when using restored vmstate.
Signed-off-by: Pavel Dovgalyuk
Acked-by: Paolo Bonzini
---
include/qemu/timer.h | 14 ++
replay/replay-in
From: Pavel Dovgalyuk
This patch adds description of the replay log file format
into the docs/replay.txt.
Signed-off-by: Pavel Dovgalyuk
Acked-by: Paolo Bonzini
---
docs/replay.txt | 69 +++
1 file changed, 69 insertions(+)
diff --git a/d
From: Alex Bennée
This is modelled after the iothread mutex lock. We keep a TLS flag to
indicate when that thread has acquired the lock and assert we don't
double-lock or release when we shouldn't have.
Signed-off-by: Alex Bennée
Tested-by: Pavel Dovgalyuk
---
replay/replay-internal.c | 11
This patch does not allows saving/loading vmstate when
replay events queue is not empty. There is no reliable
way to save events queue, because it describes internal
coroutine state. Therefore saving and loading operations
should be deferred to another record/replay step.
Signed-off-by: Pavel Dovg
This patch fixes assignment to internal events_enabled variable.
Now it is set only in record/replay mode. This affects the behavior
of the external functions that check this flag.
Signed-off-by: Pavel Dovgalyuk
Acked-by: Paolo Bonzini
---
replay/replay-events.c |8 +---
1 file changed,
From: Pavel Dovgalyuk
This patch disables setting '-snapshot' option on by default
in record/replay mode. This is needed for creating vmstates in record
and replay modes.
Signed-off-by: Pavel Dovgalyuk
---
vl.c |8 +++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/vl.c
This patch adds saving and restoring of the icount warp
timers in the vmstate.
It is needed because there timers affect the virtual clock value.
Therefore determinism of the execution in icount record/replay mode
depends on determinism of the timers.
Signed-off-by: Pavel Dovgalyuk
Acked-by: Paolo
From: Pavel Dovgalyuk
This patch enables making snapshots with blkreplay used in
block devices.
This function is required to make bdrv_snapshot_goto without
calling .bdrv_open which is not implemented.
Signed-off-by: Pavel Dovgalyuk
---
block/blkreplay.c |8
1 file changed, 8 inse
From: Pavel Dovgalyuk
This patch allows using '-snapshot' behavior in record/replay mode.
blkreplay layer creates temporary overlays on top of underlaying
disk images. It is needed, because creating an overlay over blkreplay
breaks the determinism.
This patch creates similar temporary overlay (wh
Flushing TB cache is required because TBs key in the cache may match
different code which existed in the previous state.
Signed-off-by: Pavel Dovgalyuk
Signed-off-by: Maria Klimushenkova
---
exec.c |1 +
1 file changed, 1 insertion(+)
diff --git a/exec.c b/exec.c
index 4722e52..ff31e71 100
Asynchronous events saved at checkpoints may invoke
callbacks when processed. These callbacks may also generate/read
new events (e.g. clock reads). Therefore event processing flag must be
reset before callback invocation.
Signed-off-by: Pavel Dovgalyuk
Acked-by: Paolo Bonzini
---
replay/replay-
HPET saves its state by calculating the current time and recovers timer
offset using this calculated value. But these calculations include
divisions and multiplications. Therefore the timer state cannot be recovered
precise enough.
This patch introduces saving of the original value of the offset to
It is needed when exception_index is already set to some meaningful value.
Signed-off-by: Pavel Dovgalyuk
Signed-off-by: Paolo Bonzini
---
accel/tcg/cpu-exec.c |5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c
index 280200f..9
Accidentally sent draft version of patches to the mailing list.
Please consider this one as the correct one.
This set of patches includex fixes from Alex Bennée for fixing
BQL and replay locks after inventing the MTTCG. It also includes some
additional replay patches that makes this set of fixes w
Le 11/01/2018 à 01:17, Samuel Thibault a écrit :
> Signed-off-by: Samuel Thibault
>
> ---
> Difference from v1: handle failure of put_user_u32 with goto efault;
> ---
> linux-user/syscall.c | 16
> 1 file changed, 16 insertions(+)
>
> diff --git a/linux-user/syscall.c b/linux-u
On Wed, Jan 10, 2018 at 06:54:45AM -0600, Eric Blake wrote:
> On 01/10/2018 02:26 AM, Peter Xu wrote:
>
> >> The later initialization of the monitor_lock mutex is a potential
> >> semantic change. Please beef up the commit message to document why it
> >> is safe. In fact, I requested this back o
On Wed, Jan 10, 2018 at 06:59:13AM -0600, Eric Blake wrote:
> On 01/10/2018 01:57 AM, Peter Xu wrote:
> > On Tue, Jan 09, 2018 at 04:53:40PM -0600, Eric Blake wrote:
> >> On 12/19/2017 02:45 AM, Peter Xu wrote:
> >>> We can simplify object_property_get_str() using the new
> >>> qobject_get_try_str(
#ifdef CONFIG_USER_ONLY
int riscv_cpu_mmu_index(CPURISCVState *env, bool ifetch)
{
return 0;
}
bool riscv_cpu_exec_interrupt(CPUState *cs, int interrupt_request)
{
return false;
}
int riscv_cpu_handle_mmu_fault(CPUState *cs, vaddr address,
int access_type, int mmu_idx)
{
cs->e
On 11/01/18 04:08, Alexey Kardashevskiy wrote:
As stated in the 1ad9f0a464fe commit log, the returned entries are not
a while PTEG. It was not a problem before 1ad9f0a464fe as it would read
s/while/whole/?
a single record assuming it contains a whole PTEG but now the code tries
reading the e
301 - 377 of 377 matches
Mail list logo
