On Mon, Nov 09, 2015 at 07:46:55PM +1100, David Gibson wrote:
> On Mon, Nov 09, 2015 at 03:24:15PM +1100, David Gibson wrote:
> > On Tue, Nov 03, 2015 at 03:38:19PM +0530, Bharata B Rao wrote:
> > > KVM_PPC_ALLOCATE_HTAB ioctl can return -ENOMEM for KVM guests and QEMU
> > > never handled this
On Mon, 11/09 11:42, Stefan Hajnoczi wrote:
> On Mon, Nov 09, 2015 at 10:56:39AM +0800, Fam Zheng wrote:
> > v3: Don't reuse coroutine in bdrv_aio_ioctl. [Stefan]
> > Recursely call .bdrv_drain callback only. [Stefan, Paolo]
>
> I don't understand this change. I thought you want
On Fri, 6 Nov 2015 16:02:49 +0800
Jason Wang wrote:
> Guest always get zero when reading queue_enable. This violates
> spec. Fixing this by setting the queue_enable to true during any guest
> writing and setting it to zero during reset.
>
> Cc: Michael S. Tsirkin
On 9 November 2015 at 12:51, Peter Maydell wrote:
> On 9 November 2015 at 10:08, Stefan Hajnoczi wrote:
>> The following changes since commit c4a7bf54e588ff2de9fba0898b686156251b2063:
>>
>> Merge remote-tracking branch
On 11/09/2015 02:59 AM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> On 11/06/2015 09:03 AM, Markus Armbruster wrote:
>>> Eric Blake writes:
> [...]
Hopefully, we are converging on something that will be ready
for a pull request, especially
Sorry again for the late review, here are my comments:
On Fri 16 Oct 2015 10:57:45 AM CEST, Wen Congyang wrote:
> +void qmp_x_blockdev_change(ChangeOperation op, const char *parent,
> + bool has_child, const char *child,
> + bool has_new_node,
Signed integer overflow in C is undefined behaviour, and the compiler
is at liberty to assume it can never happen and optimize accordingly.
In particular, the subtractions in hpet_time_after() and hpet_time_after64()
were causing OSX clang to optimize the code such that it was prone to
hangs and
Eric Blake writes:
> Consolidate two common sequences of clash detection into a
> new QAPISchemaObjectType.check_clash() helper method.
>
> No change to generated code.
>
> Signed-off-by: Eric Blake
>
> ---
> v10: rebase on new Variants.check_clash()
> v9:
This patch enables the migration of the entire array of MAC registers
during live migration. The entire array is just 128 KB long, so
practically no penalty should be felt when transmitting it, additionally
to the previously transmitted individual registers. The advantage here is
eliminating the
Previously, these 64-bit registers did not stick at their maximal
values when (and if) they reached them, as they should do, according to
the specs.
This patch introduces a function that takes care of such registers,
avoiding code duplication, making the relevant parts more compatible
with the
According to Intel's specs, these counters (as the other Statistic
registers) stick at 0x when this maximal value is reached.
Previously, they would reset after the max. value.
Signed-off-by: Leonid Bloch
Signed-off-by: Dmitry Fleytman
On 09/11/2015 10:03, arei.gong...@huawei.com wrote:
> From: Gonglei
>
> 1. avoid possible superflous checking
> 2. make code more robustness
>
> Signed-off-by: Gonglei
> Reviewed-by: Fam Zheng
> ---
> v3: change the third
Can you reproduce with Ceph debug logging enabled (i.e. debug rbd=20 in your
ceph.conf)? If you could attach the log to the Ceph tracker ticket I opened
[1], that would be very helpful.
[1] http://tracker.ceph.com/issues/13726
Thanks,
Jason
- Original Message -
> From: "Alexandre
Eric Blake writes:
> Future commits will migrate semantic checking away from parsing
> and over to the various QAPISchema*.check() methods. But to
> report an error message about an incorrect semantic use of a
> member of an object type, it helps to know which type, command,
On 9 November 2015 at 13:40, Paolo Bonzini wrote:
>
>
> On 09/11/2015 14:30, Peter Maydell wrote:
>> After a bunch of "try building specific object files with optimization
>> off to see where the problem goes away" tests, I've narrowed the
>> problem down further: if you tell
Eric Blake writes:
> Consolidate two common sequences of clash detection into a
> new QAPISchemaObjectType.check_clash() helper method.
>
> No change to generated code.
>
> Signed-off-by: Eric Blake
>
> ---
> v10: rebase on new Variants.check_clash()
> v9:
On Fri, Nov 06, 2015 at 09:12:34AM +, Xulei (Stone) wrote:
>
> >On Wed, Nov 04, 2015 at 08:48:20AM +0800, Gonglei wrote:
> >> On 2015/11/3 14:58, Xulei (Stone, Euler) wrote:
> >> > On qemu-kvm platform, when I reset a VM through "virsh reset", and
> >> > coincidently
> >> > the VM is in
This patch adds pkeys support for qemu migration.
Signed-off-by: Huaitong Han
---
target-i386/machine.c | 23 +++
1 file changed, 23 insertions(+)
diff --git a/target-i386/machine.c b/target-i386/machine.c
index a0df64b..1b190c7 100644
---
This series fixes issues with packet/octet counting in e1000's Statistic
registers, fixes a bug in the packet address filtering procedure, and
implements many MAC registers that were absent before, some Statistic
counters among them.
Besides this, the series introduces a parameter which, if set
Eric Blake writes:
> Checking that a given QAPISchemaObjectTypeVariant.name is a
> member of the corresponding QAPISchemaEnumType of the owning
> QAPISchemaObjectTypeVariants.tag_member ensures that there are
> no collisions in the generated C union for those tag values
>
On Wed 04 Nov 2015 07:57:33 PM CET, Max Reitz wrote:
> All error paths after a successful bdrv_open() of target_bs should
> contain a bdrv_unref(target_bs). This one did not yet, so add it.
>
> Signed-off-by: Max Reitz
Reviewed-by: Alberto Garcia
> - Original Message -
>> From: "Alexandre DERUMIER"
>> To: "ceph-devel"
>> Cc: "qemu-devel" , jdur...@redhat.com
>> Sent: Monday, November 9, 2015 5:48:45 AM
>> Subject: Re: [Qemu-devel] qemu : rbd block driver
On Fri, 06 Nov 2015 18:45:04 +0100
Greg Kurz wrote:
> When running a fully emulated device in cross-endian conditions, including
> a virtio 1.0 device offered to a big endian guest, we need to fix the vnet
> headers. This is currently handled by the
Previously, if promiscuous unicast was enabled, a packet was received
straight away, even if it was a multicast or a broadcast packet. This
patch fixes that behavior, while making the filtering procedure a bit
more human-readable.
Signed-off-by: Leonid Bloch
On 09/11/2015 13:01, Pierre Morel wrote:
>>> This leads to have UINT64_MAX represented with {1, 0} instead of
>>> {0, UINT64_MAX} while {1, 0} is 2^64. This again leads to have
>>> unnecessary and obfuscating transformations with int128_2_64() to
>>> test for UINT64_MAX and return {1,0} in
Eric Blake writes:
> From: Markus Armbruster
>
> Reduce the ugly flat union / simple union conditional by doing just
> the essential work here, namely setting self.tag_member.
> Move the rest to callers.
>
> Signed-off-by: Markus Armbruster
* Bharata B Rao (bhar...@linux.vnet.ibm.com) wrote:
> On Mon, Nov 09, 2015 at 11:03:02AM +, Dr. David Alan Gilbert wrote:
> >
> >
> > Oh, I think I see it; the following is untested, I'll grab a machine to
> > try it on but it will take a little while:
> >
> > diff --git a/migration/ram.c
On Mon, Nov 09, 2015 at 05:03:30PM +0800, arei.gong...@huawei.com wrote:
> From: Gonglei
>
> 1. avoid possible superflous checking
> 2. make code more robustness
>
> Signed-off-by: Gonglei
> Reviewed-by: Fam Zheng
> ---
> v3:
>>Can you reproduce with Ceph debug logging enabled (i.e. debug rbd=20 in your
>>ceph.conf)? If you could attach the log to the Ceph tracker ticket I opened
>>[1], that would be very helpful.
>>
>>[1] http://tracker.ceph.com/issues/13726
yes,I'm able to reproduce it 100%, I have attached the
This patch adds pkeys support for qemu cpuid handling.
Signed-off-by: Huaitong Han
---
target-i386/cpu.c | 21 -
target-i386/cpu.h | 4
2 files changed, 24 insertions(+), 1 deletion(-)
diff --git a/target-i386/cpu.c b/target-i386/cpu.c
index
This patch adds pkeys support for qemu xsave state handling.
Signed-off-by: Huaitong Han
---
target-i386/cpu.c | 2 ++
target-i386/cpu.h | 3 +++
target-i386/kvm.c | 3 +++
3 files changed, 8 insertions(+)
diff --git a/target-i386/cpu.c b/target-i386/cpu.c
index
The protection-key feature provides an additional mechanism by which IA-32e
paging controls access to usermode addresses.
Hardware support for protection keys for user pages is enumerated with CPUID
feature flag CPUID.7.0.ECX[3]:PKU. Software support is CPUID.7.0.ECX[4]:OSPKE
with the setting of
On Wed 04 Nov 2015 07:57:44 PM CET, Max Reitz wrote:
> @@ -3519,11 +3537,18 @@ void qmp_x_blockdev_del(bool has_id, const char *id,
> bdrv_get_device_or_node_name(bs));
> goto out;
> }
> +
> +if (!blk && !bs->monitor_list.tqe_prev) {
> +
On 09/11/2015 12:55, Huaitong Han wrote:
> @@ -351,6 +362,7 @@ static const char *cpuid_6_feature_name[] = {
>CPUID_7_0_EBX_FSGSBASE, CPUID_7_0_EBX_HLE, CPUID_7_0_EBX_AVX2,
>CPUID_7_0_EBX_ERMS, CPUID_7_0_EBX_INVPCID, CPUID_7_0_EBX_RTM,
>CPUID_7_0_EBX_RDSEED */
On 9 November 2015 at 10:08, Stefan Hajnoczi wrote:
> The following changes since commit c4a7bf54e588ff2de9fba0898b686156251b2063:
>
> Merge remote-tracking branch 'remotes/jnsnow/tags/ide-pull-request' into
> staging (2015-11-07 19:55:15 +)
>
> are available in the
This still affects the distro proxmox 4.0 and many users are affected
"http://forum.proxmox.com/threads/24362-PCIe-passthrough-does-not-work;
, is there any detail you guys need to fix this?
--
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to
The array of uint8_t's which is introduced here, contains useful metadata
about the MAC registers: if a register should be always accessible, or if
it is accessible, but partly implemented, or if the register requires a
certain compatibility flag to be accessed. Currently, 5 hypothetical flags
are
These registers appear in Intel's specs, but were not implemented.
These registers are now implemented trivially, i.e. they are initiated
with zero values, and if they are RW, they can be written or read by the
driver, or read only if they are R (essentially retaining their zero
values). For these
Hi
- Original Message -
> On 9 November 2015 at 12:51, Peter Maydell wrote:
> > On 9 November 2015 at 10:08, Stefan Hajnoczi wrote:
> >> The following changes since commit
> >> c4a7bf54e588ff2de9fba0898b686156251b2063:
> >>
> >> Merge
On Mon, Nov 09, 2015 at 11:03:02AM +, Dr. David Alan Gilbert wrote:
>
>
> Oh, I think I see it; the following is untested, I'll grab a machine to
> try it on but it will take a little while:
>
> diff --git a/migration/ram.c b/migration/ram.c
> index 62cf42b..e6db965 100644
> ---
On Mon, 9 Nov 2015 13:35:51 +0200
"Michael S. Tsirkin" wrote:
> On Fri, Oct 30, 2015 at 01:55:59PM +0800, Xiao Guangrong wrote:
> > Implement ObjectType which is used by NVDIMM _DSM method in
> > later patch
> >
> > Signed-off-by: Xiao Guangrong
Eric Blake writes:
> Right now, our ad hoc parser ensures that we cannot have a
> flat union that introduces any qapi member names that would
> conflict with the non-variant qapi members already present
> from the union's base type (see flat-union-clash-member.json).
> We want
On Wed 04 Nov 2015 07:57:42 PM CET, Max Reitz wrote:
> There are no users of bdrv_close() left, except for one of bdrv_open()'s
> failure paths, bdrv_close_all() and bdrv_delete(), and that is good.
> Make bdrv_close() static so nobody makes the mistake of directly using
> bdrv_close() again.
>
>
On Wed 04 Nov 2015 07:57:34 PM CET, Max Reitz wrote:
> If block_job_create() fails, it should release its reference to the
> job's BDS. Normally, this is done in the callback provided by the
> caller, but that callback will not be invoked if the block job failed to
> be created.
>
> Signed-off-by:
On 09/11/2015 14:30, Peter Maydell wrote:
> After a bunch of "try building specific object files with optimization
> off to see where the problem goes away" tests, I've narrowed the
> problem down further: if you tell clang to disable optimization by
> adding __attribute__ ((optnone)) to the two
On Fri, 6 Nov 2015 16:02:44 +0800
Jason Wang wrote:
> We don't migrate the followings fields for virtio-pci:
>
> uint32_t dfselect;
> uint32_t gfselect;
> uint32_t guest_features[2];
> struct {
> uint16_t num;
> bool enabled;
> uint32_t desc[2];
> uint32_t
On 9 November 2015 at 11:46, Peter Maydell wrote:
> Yep, just confirmed that current master with commit b553a0428
> reverted displays the bug again.
After a bunch of "try building specific object files with optimization
off to see where the problem goes away" tests,
This fixes some alignment and cosmetic issues. The changes are made
in order that the following patches in this series will look like
integral parts of the code surrounding them, while conforming to the
coding style. Although some changes in unrelated areas are also made.
Signed-off-by: Leonid
This implements the following Statistic registers (various counters)
according to Intel's specs:
TSCTC GOTCL GOTCH GORCL GORCH MPRC BPRC RUCROC
BPTC MPTC PTC... PRC...
Signed-off-by: Leonid Bloch
Signed-off-by: Dmitry Fleytman
Eric Blake writes:
> From: Markus Armbruster
>
> QAPISchemaObjectTypeMember.check() currently does four things:
>
> 1. Compute self.type
>
> 2. Accumulate members in all_members
>
>Only one caller cares: QAPISchemaObjectType.check() uses it to
>
On 11/09/2015 05:31 AM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> From: Markus Armbruster
>>
>> QAPISchemaObjectTypeMember.check() currently does four things:
>>
>> 4. Check for collisions
>>
>>This piggyback on 3: before adding a new entry,
On Fri 16 Oct 2015 10:57:46 AM CEST, Wen Congyang wrote:
> +.name = "blockdev_change",
> +.args_type = "op:s,parent:B,child:B?,node:?",
> +.params = "operation parent [child] [node]",
[...]
> +/*
> + * FIXME: we must specify the parameter child,
On 09/11/15 19:05, John Snow wrote:
> Marc noticed that a recent ATAPI permissions fix broke NetBSD 7.0's
> installer ISO.
>
> The problem is that it's meaningless to check for !(cmd->flags & nondata)
> if the command isn't supported, since all unsupported commands have
> _no_ flags.
On Mon, Nov 02, 2015 at 11:53:26PM +0100, Andrew Jones wrote:
> cacheflush is an arm-specific syscall that qemu built for arm
> uses. Add it to the whitelist, but only if we're linking with
> a recent enough libseccomp.
>
> Signed-off-by: Andrew Jones
> ---
> v3: deal with
On 11/06/2015 02:13 PM, Denis V. Lunev wrote:
>> That is a case of using libvirt to trigger internal snapshots...
>>
>>> The HMP monitor is legacy and also not used by modern libvirt.
>> ...and libvirt is forced to use HMP for internal snapshots, since we
>> _still_ haven't exposed internal
[adding Markus for a qapi question]
On 11/09/2015 10:28 AM, Juan Quintela wrote:
> From: "Dr. David Alan Gilbert"
>
> Once postcopy is enabled (with migrate_set_capability), the migration
> will still start on precopy mode. To cause a transition into postcopy
> the:
>
>
For append file open modes, use FILE_APPEND_DATA for the desired access for
writing at the end of the file.
Signed-off-by: Kirk Allan
---
qga/commands-win32.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/qga/commands-win32.c
On 11/09/2015 08:56 PM, Cornelia Huck wrote:
> On Fri, 6 Nov 2015 16:02:44 +0800
> Jason Wang wrote:
>
>> We don't migrate the followings fields for virtio-pci:
>>
>> uint32_t dfselect;
>> uint32_t gfselect;
>> uint32_t guest_features[2];
>> struct {
>> uint16_t num;
buffer_find_nonzero_offset() is a hot function during live migration.
Now it use SSE2 intructions for optimization. For platform supports
AVX2 instructions, use the AVX2 instructions for optimization can help
to improve the performance about 30% comparing to SSE2.
Zero page check can be faster
Add the '--enable-avx2' & '--disable-avx2' option so as to config
the AVX2 instruction optimization.
By default, avx2 optimization is enabled, if '--disable-avx2' is not
set, configure will detect if the compiler can support AVX2 option,
if yes, AVX2 optimization is eabled, else disabled.
buffer_find_nonzero_offset() is a hot function during live migration.
Now it use SSE2 intructions for optimization. For platform supports
AVX2 instructions, use the AVX2 instructions for optimization can help
to improve the performance about 30% comparing to SSE2.
Zero page check can be faster
On 11/10/2015 02:57 PM, Sukadev Bhattiprolu wrote:
Alexey Kardashevskiy [a...@ozlabs.ru] wrote:
| On 11/05/2015 10:06 AM, Sukadev Bhattiprolu wrote:
| >Implement RTAS_SYSPARM_PROCESSOR_MODULE_INFO parameter to rtas_get_sysparm()
| >call in qemu. This call returns the processor module (socket),
Alexey Kardashevskiy [a...@ozlabs.ru] wrote:
| >| When exactly does a socket become a module? The SPAPR spec uses "sockets"
here.
| >
| >I am trying to get the terminology too :-) Is socket a slot where a
| >module is attached?
|
| Sorry, no idea.
Ok.
|
|
| >
| >I will change the variable
On 11/09/2015 10:59 PM, Leonid Bloch wrote:
> The array of uint8_t's which is introduced here, contains useful metadata
> about the MAC registers: if a register should be always accessible, or if
> it is accessible, but partly implemented, or if the register requires a
> certain compatibility
On 11/09/2015 08:44 PM, Max Reitz wrote:
> Just specifying a custom string is simpler in basically all places that
> used it, and in addition, specifying the BB or node name is something we
> generally do not do in other error messages when opening a BDS, so we
> should not do it here.
>
> This
On Tue, Nov 10, 2015 at 04:49:12AM +0100, Max Reitz wrote:
> On 10.11.2015 04:45, Eric Blake wrote:
> > [side question]
> >
> > On 11/09/2015 08:27 PM, Max Reitz wrote:
> >> Overall, this series adds more uses for BlockBackends and makes the code
> >> follow the "reference theory" more closely,
On 11/09/2015 12:11 PM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> On 11/09/2015 06:00 AM, Markus Armbruster wrote:
>>> Eric Blake writes:
>>>
Consolidate two common sequences of clash detection into a
new
Alexey Kardashevskiy [a...@ozlabs.ru] wrote:
| On 11/05/2015 10:06 AM, Sukadev Bhattiprolu wrote:
| >Implement RTAS_SYSPARM_PROCESSOR_MODULE_INFO parameter to rtas_get_sysparm()
| >call in qemu. This call returns the processor module (socket), chip and core
| >information as specified in section
David Gibson [da...@gibson.dropbear.id.au] wrote:
| On Wed, Nov 04, 2015 at 03:06:05PM -0800, Sukadev Bhattiprolu wrote:
| > Implement RTAS_SYSPARM_PROCESSOR_MODULE_INFO parameter to rtas_get_sysparm()
| > call in qemu. This call returns the processor module (socket), chip and core
| > information
> Rather than trying to cater to multiple assembly instruction implementations
> ourselves, have you tried taking the ideas in this earlier thread?
> https://lists.gnu.org/archive/html/qemu-devel/2015-10/msg05298.html
>
> Ideally, libc's memcmp() will already be using the most efficient assembly
On 11/10/2015 12:04 AM, Kevin Wolf wrote:
> Am 16.10.2015 um 10:57 hat Wen Congyang geschrieben:
>> +##
>> +# @ChangeOperation:
>> +#
>> +# An enumeration of block device change operation.
>> +#
>> +# @add: Add a new block driver state to a existed block driver state.
>> +#
>> +# @delete: Delete a
On 11/09/2015 09:14 PM, Cornelia Huck wrote:
> On Fri, 6 Nov 2015 16:02:49 +0800
> Jason Wang wrote:
>
>> Guest always get zero when reading queue_enable. This violates
>> spec. Fixing this by setting the queue_enable to true during any guest
>> writing and setting it to
The monitor does hold references to some BlockBackends so it should have
a list of those BBs; blk_backends is a different list, as it contains
references to all BBs (after a follow-up patch, that is), and that
should not be changed because we do need such a list.
monitor_remove_blk() is
This function skips to the next BlockBackend for which blk_is_inserted()
is true.
Signed-off-by: Max Reitz
Reviewed-by: Eric Blake
---
block/block-backend.c | 15 +++
include/sysemu/block-backend.h | 1 +
2 files changed, 16
These functions will be changed to iterate through the BDS trees as
defined by the BlockBackends instead of the list of root BDS, therefore
prepare moving their code to the BlockBackend level.
Signed-off-by: Max Reitz
Reviewed-by: Eric Blake
---
There may be BlockBackends which are not returned by blk_by_name(), but
do exist and have a name. blk_name_taken() allows testing whether a
specific name is in use already, independent of whether the BlockBackend
with that name is accessible through blk_by_name().
Signed-off-by: Max Reitz
There are no callers to bdrv_open() or bdrv_open_inherit() left that
pass a pointer to a non-NULL BDS pointer as the first argument of these
functions, so we can finally drop that parameter and just make them
return the new BDS.
Generally, the following pattern is applied:
bs = NULL;
ret
It is unused now, so we may just as well drop it.
Signed-off-by: Max Reitz
---
block.c | 5 -
include/block/block.h | 1 -
2 files changed, 6 deletions(-)
diff --git a/block.c b/block.c
index 395b7b0..0531992 100644
--- a/block.c
+++ b/block.c
@@ -243,11
On 10.11.2015 04:45, Eric Blake wrote:
> [side question]
>
> On 11/09/2015 08:27 PM, Max Reitz wrote:
>> Overall, this series adds more uses for BlockBackends and makes the code
>> follow the "reference theory" more closely, in that any BlockBackend
>> created (through -drive or blockdev-add) has
Its only caller is blk_new_open(), so we can just inline it there. Since
bdrv_new_root() is only a wrapper around bdrv_new(), we can just use
bdrv_new() instead.
Signed-off-by: Max Reitz
---
block/block-backend.c | 29 +++--
1 file changed, 7
[side question]
On 11/09/2015 08:27 PM, Max Reitz wrote:
> Overall, this series adds more uses for BlockBackends and makes the code
> follow the "reference theory" more closely, in that any BlockBackend
> created (through -drive or blockdev-add) has a reference count of 1, and
> that reference
On 11/09/2015 05:38 AM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> From: Markus Armbruster
>>
>> Reduce the ugly flat union / simple union conditional by doing just
>> the essential work here, namely setting self.tag_member.
>> Move the rest to
On 11/09/2015 05:56 AM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> Right now, our ad hoc parser ensures that we cannot have a
>> flat union that introduces any qapi member names that would
>> conflict with the non-variant qapi members already present
>> from the
KVM_PPC_ALLOCATE_HTAB ioctl can return -ENOMEM for KVM guests and QEMU
never handled this correctly. But this didn't cause any problems till
now as KVM_PPC_ALLOCATE_HTAB ioctl returned with smaller than requested
HTAB when enough contiguous memory wasn't available in the host.
After the proposed
KVM_PPC_ALLOCATE_HTAB ioctl can now return errors and handle that
correctly in QEMU.
Also fix !CONFIG_KVM implementation of kvmppc_reset_htab() to return
0 instead of -1.
Changes in v1
- Add a patch to fix !CONFIG_KVM implementation of kvmppc_reset_htab()
so that make check passes
The !CONFIG_KVM implementation of kvmppc_reset_htab() returns -1
by default. Change this to return 0 so that we fall back to user space
HTAB allocation for emulated guests.
This fixes the make check failures for ppc64 emulated target.
Signed-off-by: Bharata B Rao
---
On 11/09/2015 06:07 AM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> Checking that a given QAPISchemaObjectTypeVariant.name is a
>> member of the corresponding QAPISchemaEnumType of the owning
>> QAPISchemaObjectTypeVariants.tag_member ensures that there are
>> no
The information which BB is concerned does not seem useful enough to
justify its existence in most other place (which may be related to qemu
printing the -drive parameter in question anyway, and for blockdev-add
the attribution is naturally unambiguous). Furthermore, as of a future
patch,
After a lot has been restructed in the block layer in the past, we can
now reap at least one of the fruits: Make bdrv_open() return a BDS!
I tried to squeeze patch 8 of this series into my bdrv_close_all()
series but noticed that won't work because bdrv_open_inherit() still has
one instance of
The only caller of bdrv_close() left is bdrv_delete(). We may as well
assert that, in a way (there are some things in bdrv_close() that make
more sense under that assumption, such as the call to
bdrv_release_all_dirty_bitmaps() which in turn assumes that no frozen
bitmaps are attached to the BDS).
If bdrv_open_inherit() creates a snapshot BDS and *pbs is NULL, that
snapshot BDS should be returned instead of the BDS under it.
To this end, bdrv_append_temp_snapshot() now returns the snapshot BDS
instead of just appending it on top of the snapshotted BDS, and that
function is made static
Just specifying a custom string is simpler in basically all places that
used it, and in addition, specifying the BB or node name is something we
generally do not do in other error messages when opening a BDS, so we
should not do it here.
This changes the output for iotest 036 (to the better, in
On 11/09/2015 07:51 PM, Liang Li wrote:
> buffer_find_nonzero_offset() is a hot function during live migration.
> Now it use SSE2 intructions for optimization. For platform supports
> AVX2 instructions, use the AVX2 instructions for optimization can help
> to improve the performance about 30%
On 11/09/2015 08:13 AM, Markus Armbruster wrote:
> Markus Armbruster writes:
>
>> Eric Blake writes:
>>
>>> Right now, our ad hoc parser ensures that we cannot have a
>>> flat union that introduces any qapi member names that would
>>> conflict with the
On 11/09/2015 07:49 AM, Markus Armbruster wrote:
> Eric Blake writes:
>
>> Consolidate two common sequences of clash detection into a
>> new QAPISchemaObjectType.check_clash() helper method.
>>
>> No change to generated code.
>>
>> Signed-off-by: Eric Blake
buffer_find_nonzero_offset() is a hot function during live migration.
Now it use SSE2 intructions for optimization. For platform supports
AVX2 instructions, use the AVX2 instructions for optimization can help
to improve the performance about 30% comparing to SSE2.
Zero page check can be faster
Add the '--enable-avx2' & '--disable-avx2' option so as to config
the AVX2 instruction optimization.
By default, avx2 optimization is enabled, if '--disable-avx2' is not
set, configure will detect if the compiler can support AVX2 option,
if yes, AVX2 optimization is eabled, else disabled.
buffer_find_nonzero_offset() is a hot function during live migration.
Now it use SSE2 intructions for optimization. For platform supports
AVX2 instructions, use the AVX2 instructions for optimization can help
to improve the performance about 30% comparing to SSE2.
Zero page check can be faster
On Sat, Nov 07, 2015 at 06:54:55PM +0300, Denis V. Lunev wrote:
This:
> +int bdrv_all_find_snapshot(const char *name, BlockDriverState **first_bad_bs)
> +{
> +QEMUSnapshotInfo sn;
> +int err = 0;
> +BlockDriverState *bs = NULL;
> +
> +while (err == 0 && (bs = bdrv_next(bs))) {
>
From: "Dr. David Alan Gilbert"
'file' becomes confusing when you have flows in each direction;
rename to make it clear.
This leaves just the main forward direction ms->file, which is used
in a lot of places and is probably not worth renaming given the churn.
Signed-off-by:
1 - 100 of 465 matches
Mail list logo