On Wed, 12/03 22:28, Bryan D. Payne wrote:
> >
> > It is critical as a transport data structure. You have to define a
> > byte-by-byte
> > layout (concerning endianness and padding) and use padding fields together
> > with
> > QEMU_PACKED, so the representation is not dependent on alignment
>
>
>
Hi,
Recently, my colleagues encountered a confused issue about ipxe in qemu
release version, both qemu-2.1 and qemu-2.2-rc4.
Details:
When vm configured a lot of network devices, ipxe install failed.
The result as follow:
Configuring (net0 00:e0:fc::00:fc:54)Error 0x04
On 03.12.2014 18:25, Max Reitz wrote:
On 2014-12-02 at 15:33, Peter Lieven wrote:
this patch finally introduce multiread support to virtio-blk while
multiwrite support was there for a long time read support was missing.
To achieve this the patch does serveral things which might need futher
expl
On 2014/12/4 14:51, Amos Kong wrote:
> 'local_err' was defined for collecting error insider qemu_opts_absorb_qdict().
> We should set error to 'errp', then the error can be passed to caller,
> and free the memory in future.
>
> Reviewed-by: Amos Kong
Thanks :)
Regards,
-Gonglei
On Thu, Dec 04, 2014 at 02:34:11PM +0800, arei.gong...@huawei.com wrote:
> From: Gonglei
>
> Variable local_err going out of scope
> leaks the storage it points to.
>
> Cc: Markus Armbruster
> Signed-off-by: Gonglei
> ---
> block/rbd.c | 8
> 1 file changed, 4 insertions(+), 4 deleti
On Thu, Dec 04, 2014 at 02:18:16PM +0800, Fam Zheng wrote:
> Using "has" is more slick because which(1) is not always there.
>
> Signed-off-by: Fam Zheng
> ---
> configure | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/configure b/configure
> index 47048f0..cae588c 100
From: Gonglei
Variable local_err going out of scope
leaks the storage it points to.
Cc: Markus Armbruster
Signed-off-by: Gonglei
---
block/rbd.c | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/block/rbd.c b/block/rbd.c
index 5b5a64a..f3ab2dd 100644
--- a/block/rbd.
Am 04.12.2014 um 07:18 schrieb Fam Zheng:
> Using "has" is more slick because which(1) is not always there.
>
> Signed-off-by: Fam Zheng
> ---
> configure | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/configure b/configure
> index 47048f0..cae588c 100755
> --- a/confi
On Wed, 3 Dec 2014 15:38:27 +0100
Jens Freimann wrote:
> Cornelia, Alex, Christian,
>
> here are some patches targeted at QEMU 2.3.
>
> * Patches 1-3 are simple clean-ups
> * Patch 4 moves some register values into struct kvm_run,
> thereby saving us a lot of set_one_reg triggered syscall
>
> It is critical as a transport data structure. You have to define a
> byte-by-byte
> layout (concerning endianness and padding) and use padding fields together
> with
> QEMU_PACKED, so the representation is not dependent on alignment
This makes sense for network protocols. But, in this case,
Am 03.12.2014 um 17:50 schrieb Max Reitz:
> On 2014-12-03 at 16:30, Jeff Cody wrote:
>> This removes the unneeded BLOCK_OPT_NOCOW options from vdi
>> and vpc.
>>
>> Jeff Cody (2):
>>block: remove BLOCK_OPT_NOCOW from vdi_create_opts
>>block: remove BLOCK_OPT_NOCOW from vpc_create_opts
>>
>>
Using "has" is more slick because which(1) is not always there.
Signed-off-by: Fam Zheng
---
configure | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/configure b/configure
index 47048f0..cae588c 100755
--- a/configure
+++ b/configure
@@ -2727,7 +2727,7 @@ fi
if test "$modul
On Wed, 12/03 19:37, Bryan D. Payne wrote:
> >
> > Out of curiosity, what are existing solutions?
> >
>
> Basically just attaching gdb and pulling memory out manually (or writing a
> program to do the same).
>
>
> > +struct request {
> > > +uint8_t type; /* 0 quit, 1 read, 2 write, ...
Signed-off-by: Fam Zheng
---
tests/Makefile| 2 +
tests/test-poll.c | 272 ++
2 files changed, 274 insertions(+)
create mode 100644 tests/test-poll.c
diff --git a/tests/Makefile b/tests/Makefile
index 16f0e4c..79f399d 100644
--- a/tests/
qemu_poll_set_fds + qemu_poll does the same, and when epoll is
available, it is faster.
Signed-off-by: Fam Zheng
---
include/qemu/timer.h | 13 -
main-loop.c | 35 ++-
qemu-timer.c | 28
3 files changed, 30
In qemu_poll_timerfd, we arm the timerfd with timeout_ns. The timerfd is
also watched by epollfd, so that when there is no other events,
epoll_wait will still return on time, even though we pass -1 (wait
infinitely).
Signed-off-by: Fam Zheng
---
poll-linux.c | 84
The AIO handler list is only modified by aio_set_fd_handler, so we can
easily add del poll fd there. Initialize a QEMUPoll and keep track of
all the fds, so we don't need to rebuild a GPollFD array for g_poll in
aio_poll.
Signed-off-by: Fam Zheng
---
aio-posix.c | 52
This implements qemu_poll with ppoll + epoll. The only complex part is
qemu_poll_set_fds, which will sync up epollfd with epoll_ctl by
computing the symmetric difference of previous and new fds.
The ppoll is used to retain ns precision of timeout.
Signed-off-by: Fam Zheng
---
Makefile.objs |
This is abstract of underlying poll implementation. A glib
implementation is included.
Signed-off-by: Fam Zheng
---
Makefile.objs | 2 +-
include/qemu/poll.h | 40 +++
include/qemu/typedefs.h | 2 +
poll-glib.c | 130
v2: Emulate nanoseconds precison of timeout with ppoll and timerfd.
Their performance is on par with each other, but both much better than
qemu.git:
syscall high # of fd low # of fd
-
qemu.git(ppoll) 44
>
> Out of curiosity, what are existing solutions?
>
Basically just attaching gdb and pulling memory out manually (or writing a
program to do the same).
> +struct request {
> > +uint8_t type; /* 0 quit, 1 read, 2 write, ... rest reserved */
> > +uint64_t address; /* address to read
On 2014/12/3 20:10, Gerd Hoffmann wrote:
> Hi,
>
>> Hi, Daniel
>> Actually, my team had received the requirement of changing VNC keyboard
>> layout dynamically on the scenario of Desktop Cloud. The clientele just use
>> the simplest tight vnc client, but not GTK-VNC etc. I think we should supp
This applies cases on drive-backup on blockdev-backup, except cases with
target format and mode.
Also add a case to check source == target.
Signed-off-by: Fam Zheng
---
tests/qemu-iotests/055 | 211 +
tests/qemu-iotests/055.out | 4 +-
2 files c
Also add version info for other transaction types.
Signed-off-by: Fam Zheng
---
blockdev.c | 81
qapi-schema.json | 7 +
2 files changed, 88 insertions(+)
diff --git a/blockdev.c b/blockdev.c
index f1a..a98a4f8 100644
--- a
Similar to drive-backup, but this command uses a device id as target
instead of creating/opening an image file.
Also add blocker on target bs, since the target is also a named device
now.
Add check and report error for bs == target which became possible but is
an illegal case with introduction of
v4: Rebase to block-next.
Support dataplane by acquiring AioContext. (Stefan)
v3: Address Eric's comments on documentation.
Squashed 3/4 into 2/4.
v2: Address Markus' and Eric's comments:
Fix qapi schema documentation.
Fix versioning of transactions.
Improve test case code by
On Wed, 12/03 14:57, Max Reitz wrote:
> Add a test for vmdk files which use a file with a JSON file name, and
> which then try to open extents. That should fail and the error message
> should at least try to look helpful.
>
> Signed-off-by: Max Reitz
> ---
> tests/qemu-iotests/059 | 6 ++
On Wed, 12/03 14:57, Max Reitz wrote:
> If vmdk blindly tries to use path_combine() using bs->file->filename as
> the base file name, this will result in a bad error message for JSON
> file names when calling bdrv_open(). It is better to only try
> bs->file->exact_filename; if that is empty, bs->fi
On 12/03/2014 05:37 AM, Max Reitz wrote:
> Progress may regress; this should be displayed correctly by
> qemu_progress_print().
>
> While touching that area of code, drop the redundant parentheses in the
> same condition.
>
> Signed-off-by: Max Reitz
> ---
> util/qemu-progress.c | 3 ++-
> 1 fi
On 12/03/2014 05:37 AM, Max Reitz wrote:
> Add a test for conversion between different refcount widths and errors
> specific to certain widths (i.e. snapshots with refcount_bits=1).
>
> Signed-off-by: Max Reitz
> ---
> tests/qemu-iotests/112 | 296
> +
On 12/03/2014 05:37 AM, Max Reitz wrote:
> Add a function qcow2_change_refcount_order() which allows changing the
> refcount order of a qcow2 image.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2-refcount.c | 452
> +
> block/qcow2.h |
On 12/03/2014 05:37 AM, Max Reitz wrote:
> Add a creation option to qcow2 for setting the refcount order of images
> to be created, and respect that option's value.
>
> This breaks some test outputs, fix them.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2.c | 20
> inc
On 12/03/2014 05:37 AM, Max Reitz wrote:
> Make use of qcow2_change_refcount_order() to support changing the
> refcount order with qemu-img amend.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2.c | 44 +++-
> 1 file changed, 35 insertions(+), 9 deletions
Reported-by: Markus Armbruster
Signed-off-by: Fam Zheng
Reviewed-by: Markus Armbruster
Reviewed-by: Don Koch
Reviewed-by: Max Reitz
---
block/vmdk.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/block/vmdk.c b/block/vmdk.c
index da2d323..65af414 100644
--- a/block/vmdk.c
+++ b/block/
Zeroing a buffer that will be filled right after is not necessary, and
allocating a power of two + 1 is naughty.
Suggested-by: Markus Armbruster
Signed-off-by: Fam Zheng
Reviewed-by: Don Koch
Reviewed-by: Markus Armbruster
Reviewed-by: Max Reitz
---
block/vmdk.c | 5 +++--
1 file changed, 3
It will be assigned to the return value of vmdk_read_desc.
Suggested-by: Markus Armbruster
Signed-off-by: Fam Zheng
Reviewed-by: Markus Armbruster
Reviewed-by: Don Koch
Reviewed-by: Max Reitz
---
block/vmdk.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/block/vmdk.c b/
v4: Add Don's and Max's rev-by in 1, 3, 5, 6.
2/6: Add VMFSSPARSE (Max)
4/6: Don't set errno, and add a comment.
Fam Zheng (6):
vmdk: Use g_random_int to generate CID
vmdk: Fix comment to match code of extent lines
vmdk: Clean up descriptor file reading
vmdk: Check descriptor file
commit 04d542c8b (vmdk: support vmfs files) added support of VMFS extent
type but the comment above the changed code is left out. Update the
comment so they are consistent.
Signed-off-by: Fam Zheng
---
block/vmdk.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/block/v
Since a too small file cannot be a valid VMDK image, and also since the
buffer's first 4 bytes will be unconditionally examined by
vmdk_open_sparse, let's error out the small file case to be clear.
Signed-off-by: Fam Zheng
---
block/vmdk.c | 8
1 file changed, 8 insertions(+)
diff --gi
This replaces two "time(NULL)" invocations with "g_random_int()".
According to VMDK spec, CID "is a random 32‐bit value updated the first
time the content of the virtual disk is modified after the virtual disk
is opened". Using "seconds since epoch" is just a "lame way" to generate
it, and not comp
On Wed, 12/03 15:17, Max Reitz wrote:
> On 2014-12-03 at 11:28, Fam Zheng wrote:
> >Since a too small file cannot be a valid VMDK image, and also since the
> >buffer's first 4 bytes will be unconditionally examined by
> >vmdk_open_sparse, let's error out the small file case to be clear.
> >
> >Sign
>
> In addition to Fam's review, I have a question - does this code properly
> use qemu_open() so that I can use 'add-fd' to pass in a pre-opened
> socket fd into fdset 1, then call pmemaccess with '/dev/fdset/1'? If
> not, can you please fix it to allow this usage?
>
I'm not familiar with this s
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Some tests do not work well with certain refcount widths (i.e. you
> cannot create internal snapshots with refcount_bits=1), so make those
> widths unsupported.
>
> Furthermore, add another filter to _filter_img_create in common.filter
> which filters out
The s390 instruction emulation makes GMP fail most of its tests.
I have isolated one of the problems:
How to reproduce:
gcc m.c x.s
./a.out
Correct output on actual hardware:
Incorrect output using QEMU 2.2.0 rc4:
0
File m.c:
#include
int foo();
int main() { printf("%x\n", foo()); re
On 3 December 2014 at 20:53, Chengyu Song wrote:
> The version I use is A.a Non-Confidential Beta, release on 04 September
> 2013. I don't have any later version.
We're up to rev A.d now; you can get it from
http://infocenter.arm.com/help/topic/com.arm.doc.ddi0487a.d/index.html
(if you haven't re
Hi Christopher,
> On Dec 3, 2014, at 3:25 PM, Christopher Covington wrote:
> On 12/03/2014 02:12 AM, Chengyu Song wrote:
>> In AA64 mode, certain system registers are access through MSR/MRS
>> instructions instead of MCR/MRC. This patch added more such registers:
>
> If you don't mind sharing,
On 3 December 2014 at 20:25, Christopher Covington wrote:
> Hi Chengyu,
>
> On 12/03/2014 02:12 AM, Chengyu Song wrote:
>> { .name = "PMINTENSET", .cp = 15, .crn = 9, .crm = 14, .opc1 = 0, .opc2
>> = 1,
>>.access = PL1_RW,
>>.fieldoffset = offsetof(CPUARMState, cp15.c9_pmint
Hi Chengyu,
On 12/03/2014 02:12 AM, Chengyu Song wrote:
> In AA64 mode, certain system registers are access through MSR/MRS
> instructions instead of MCR/MRC. This patch added more such registers:
If you don't mind sharing, I'm curious what motivated this patch. I have a
particular interest in h
Add setting of the CPU secure property based on the virt machine secure
property during initialization. This enables/disables secure state during
start-up.
Signed-off-by: Greg Bellows
---
hw/arm/virt.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/hw/arm/virt.c b/hw/arm/virt.
From: Fabian Aggeler
Set ARM_FEATURE_EL3 feature for CPUs that implement Security Extensions.
Signed-off-by: Fabian Aggeler
Signed-off-by: Greg Bellows
---
target-arm/cpu.c | 4
1 file changed, 4 insertions(+)
diff --git a/target-arm/cpu.c b/target-arm/cpu.c
index 0e660f9..9e8d40c 10064
When resetting a VCPU we currently call both kvm_arm_vcpu_init() and
write_kvmstate_to_list(), both of which can fail, but we never check the
return value.
The only choice here is to print an error an exit if the calls fail.
Signed-off-by: Christoffer Dall
---
target-arm/kvm32.c | 13 ++
Added a "secure" state property to the ARMCPU descriptor. This property
indicates whether the ARMCPU is enabled for secure state or not. By default it
is disabled at this time.
Signed-off-by: Greg Bellows
---
target-arm/cpu-qom.h | 2 ++
target-arm/cpu.c | 24
2 f
Add "secure" virt machine specific property to allow override of the
default secure state configuration. By default, when using the QEMU
-kernel command line argument, virt machines boot into NS/SVC. When using
the QEMU -bios command line argument, virt machines boot into S/SVC.
The secure state
Switch virt qemu machine support to use the newer object type, class, and
instance model. Added virt TypeInfo with static registration along with virt
specific class and machine structs. Also added virt class initialization
method.
Signed-off-by: Greg Bellows
---
hw/arm/virt.c | 40 +++
Add an unset_feature() function to compliment the set_feature() function. This
will be used to disable functions after they have been enabled during
initialization.
Signed-off-by: Greg Bellows
---
target-arm/cpu.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/target-arm/cpu.c b/targe
Change the Vexpress daughterboard initialization method to take a vexpress
machine state pointer instead of the daughterboard struct pointer. The machine
state now contains the daughterboard pointer.
Signed-off-by: Greg Bellows
---
hw/arm/vexpress.c | 10 +-
1 file changed, 5 insertions
Added 'secure' qemu boolean option to qemu_machine_opts[].
Signed-off-by: Greg Bellows
---
vl.c | 4
1 file changed, 4 insertions(+)
diff --git a/vl.c b/vl.c
index eb89d62..5d640f7 100644
--- a/vl.c
+++ b/vl.c
@@ -387,6 +387,10 @@ static QemuOptsList qemu_machine_opts = {
.nam
Adds base Vexpress class and machine objects and infrastructure. This is in
preparation for switching to the full QEMU object model. The base vexpress
infrastructure is intended to handle common vexpress details.
Signed-off-by: Greg Bellows
---
hw/arm/vexpress.c | 44 ++
This patchset adds functionality for enabling the ARM CPU security extensions.
At this time, the only machines supported are Versatile Express and the QEMU
ARM virtual machines both with Cortex A9 & A15.
The patchset establishes the default security state along with adding
overriding controls of t
Add "secure" Vexpress machine specific property to allow override of the
default secure state configuration. By default, when using the QEMU
-kernel command line argument, Vexpress machines boot into NS/SVC. When using
the QEMU -bios command line argument, Vexpress machines boot into S/SVC.
The
Adds setting of the CPU secure state property based on the vexpress machine
secure state property during initialization. This enablesi/disables secure
state during start-up. Changes include adding an additional secure state
boolean during vexpress CPU initialization.
Signed-off-by: Greg Bellows
Switched the Vexpress machine initialization to use the common function with
the machine pointer to board info.
Signed-off-by: Greg Bellows
---
hw/arm/vexpress.c | 26 --
1 file changed, 4 insertions(+), 22 deletions(-)
diff --git a/hw/arm/vexpress.c b/hw/arm/vexpress.c
Add Vexpress machine objects for the the Cortex A9 & A15 variants. The older
style QEMUMachine types were replaced with dedicated TypeInfo objects. The new
objects include dedicated class init functions that currently ustilze dedicated
machine init methods. The previous qemu_register_machine call
On 12/03/2014 06:37 AM, Max Reitz wrote:
> qcow2_amend_options() should not compare options against some inline
> strings but rather use the symbolic macros available for each of the
> creation options.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2.c | 31 ---
> 1
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Add a refcount_order parameter to qcow2_create2(), use that value for
> the image header and for calculating the size required for
> preallocation.
>
> For now, always pass 4.
>
> This addition requires changes to the calculation of the file size for
> t
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Add helper functions for getting and setting refcounts in a refcount
> array for any possible refcount order, and choose the correct one during
> refcount initialization.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2-refcount.c | 121
> +
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Since refcounts do not always have to be a uint16_t, all refcount blocks
> and arrays in memory should not have a specific type (thus they become
> pointers to void) and for accessing them, two helper functions are used
> (a getter and a setter). Those fun
On Wed, 19 Nov 2014, Maciej W. Rozycki wrote:
Tighten ISA level checks down to MIPS II that many of our instructions
are missing. Also make sure any 64-bit instruction enables are only
applied to 64-bit processors, that is ones that implement at least the
MIPS III ISA.
Signed-off-by: Maciej W. R
On Wed, 3 Dec 2014, Leon Alrae wrote:
> > -env->hflags |= MIPS_HFLAG_DM | MIPS_HFLAG_64 | MIPS_HFLAG_CP0;
> > +if (env->insn_flags & ISA_MIPS3)
> > +env->hflags |= MIPS_HFLAG_64;
>
> According to the CODING_STYLE braces are required even for a single
> statement "if" b
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Add a helper function for reallocating a refcount array, independent of
> the refcount order. The newly allocated space is zeroed and the function
> handles failed reallocations gracefully.
>
> The helper function will always align the buffer size to a cl
On 19/11/2014 14:20, Maciej W. Rozycki wrote:
> -env->hflags |= MIPS_HFLAG_DM | MIPS_HFLAG_64 | MIPS_HFLAG_CP0;
> +if (env->insn_flags & ISA_MIPS3)
> +env->hflags |= MIPS_HFLAG_64;
According to the CODING_STYLE braces are required even for a single
statement "if" block.
On 12/03/2014 06:37 AM, Max Reitz wrote:
> qcow2_alloc_bytes() may reuse a cluster multiple times, in which case
> the refcount is increased accordingly. However, if this would lead to an
> overflow the function should instead just not reuse this cluster and
> allocate a new one.
>
> Signed-off-by
On 12/02/2014 07:33 AM, Peter Lieven wrote:
> Signed-off-by: Peter Lieven
> ---
> block.c|2 ++
> block/accounting.c |7 +++
> block/qapi.c |2 ++
> hmp.c |6 +-
> include/block/accounting.h |3 +++
> qapi/
On 2014-12-02 at 15:33, Peter Lieven wrote:
this patch finally introduce multiread support to virtio-blk while
multiwrite support was there for a long time read support was missing.
To achieve this the patch does serveral things which might need futher
explaination:
- the whole merge and mult
On Wed, 3 Dec 2014 18:28:41 +0800
Fam Zheng wrote:
> Reported-by: Markus Armbruster
> Signed-off-by: Fam Zheng
> Reviewed-by: Markus Armbruster
> ---
Reviewed-by: Don Koch
> block/vmdk.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/block/vmdk.c b/block/vmdk.c
> index ed492b
On Wed, 3 Dec 2014 18:28:40 +0800
Fam Zheng wrote:
> It will be assigned to the return value of vmdk_read_desc.
>
> Suggested-by: Markus Armbruster
> Signed-off-by: Fam Zheng
> Reviewed-by: Markus Armbruster
> ---
Reviewed-by: Don Koch
> block/vmdk.c | 2 +-
> 1 file changed, 1 insertion(
On 12/03/2014 06:37 AM, Max Reitz wrote:
> qcow2_update_cluster_refcount() may fail, and qcow2_alloc_bytes() should
> mind that case.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2-refcount.c | 33 +
> 1 file changed, 21 insertions(+), 12 deletions(-)
>
Review
On Wed, 3 Dec 2014 18:28:38 +0800
Fam Zheng wrote:
> Zeroing a buffer that will be filled right after is not necessary, and
> allocating a power of two + 1 is naughty.
>
> Suggested-by: Markus Armbruster
> Signed-off-by: Fam Zheng
> ---
Reviewed-by: Don Koch
> block/vmdk.c | 5 +++--
> 1 f
On Wed, 3 Dec 2014 18:28:36 +0800
Fam Zheng wrote:
> This replaces two "time(NULL)" invocations with "g_random_int()".
> According to VMDK spec, CID "is a random 32‐bit value updated the first
> time the content of the virtual disk is modified after the virtual disk
> is opened". Using "seconds s
On 2014-12-03 at 16:30, Jeff Cody wrote:
This removes the unneeded BLOCK_OPT_NOCOW options from vdi
and vpc.
Jeff Cody (2):
block: remove BLOCK_OPT_NOCOW from vdi_create_opts
block: remove BLOCK_OPT_NOCOW from vpc_create_opts
block/vdi.c | 5 -
block/vpc.c | 5 -
2 files chang
On Wed, 3 Dec 2014 17:20:27 +0800
Fam Zheng wrote:
> On Wed, 12/03 09:21, Markus Armbruster wrote:
> > Fam Zheng writes:
> >
> > > Zeroing a buffer that will be filled right after is not necessary, and
> > > allocating a power of two + 1 is naughty.
> > >
> > > Suggested-by: Markus Armbruster
On 2014-12-03 at 16:30, Jeff Cody wrote:
This removes the unneeded BLOCK_OPT_NOCOW options from vdi
and vpc.
Jeff Cody (2):
block: remove BLOCK_OPT_NOCOW from vdi_create_opts
block: remove BLOCK_OPT_NOCOW from vpc_create_opts
block/vdi.c | 5 -
block/vpc.c | 5 -
2 files chang
On 2014-12-03 at 17:11, Eric Blake wrote:
On 12/03/2014 06:37 AM, Max Reitz wrote:
Refcounts may have a width of up to 64 bits, so qemu should use the same
width to represent refcount values internally.
Signed-off-by: Max Reitz
---
block/qcow2-cluster.c | 2 +-
block/qcow2-refcount.c | 42
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Refcounts may have a width of up to 64 bits, so qemu should use the same
> width to represent refcount values internally.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2-cluster.c | 2 +-
> block/qcow2-refcount.c | 42
On 12/03/2014 06:37 AM, Max Reitz wrote:
> update_refcount() and qcow2_update_cluster_refcount() currently take a
> signed addend. At least one caller passes a value directly derived from
> an absolute refcount that should be reached ("l2_refcount - 1" in
> expand_zero_clusters_in_l1()). Therefore,
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Refcounts can theoretically be of type uint64_t; in order to be able to
> represent the full range, qcow2_get_refcount() cannot use a single
> variable to represent both all refcount values and also keep some values
> reserved for errors.
>
> One solution
This removes the unneeded BLOCK_OPT_NOCOW options from vdi
and vpc.
Jeff Cody (2):
block: remove BLOCK_OPT_NOCOW from vdi_create_opts
block: remove BLOCK_OPT_NOCOW from vpc_create_opts
block/vdi.c | 5 -
block/vpc.c | 5 -
2 files changed, 10 deletions(-)
--
1.9.3
In commit fef6070, the need for NOCOW was removed from the vpc driver,
as we removed the the posix calls. However, the BLOCK_OPT_NOCOW was not
removed from vpc_create_opts. This was a mistake - remove the opt from
there as well.
Signed-off-by: Jeff Cody
---
block/vpc.c | 5 -
1 file change
In commit 7074786, the need for NOCOW was removed from the vdi driver,
as we removed the the posix calls. However, the BLOCK_OPT_NOCOW was not
removed from vdi_create_opts. This was a mistake - remove the opt from
there as well.
Signed-off-by: Jeff Cody
---
block/vdi.c | 5 -
1 file change
On 03/12/2014 16:11, Jiri Slaby wrote:
> On 10/13/2014, 03:00 PM, Paolo Bonzini wrote:
>>> +static void edu_dma_timer(void *opaque)
>>> +{
>>> + EduState *edu = opaque;
>>> + bool raise_irq = false;
>>> +
>>> + qemu_mutex_lock(&edu->dma_mutex);
>>
>> dma_mutex and mutex and irq_mutex are no
On 12/03/2014 06:37 AM, Max Reitz wrote:
> qcow2_update_cluster_refcount() does not have any quick access to the
> new refcount value, it has to call qcow2_get_refcount(). Some callers do
> not need that new value at all, others call qcow2_get_refcount()
> themselves anyway (albeit in a different c
On 10/13/2014, 03:00 PM, Paolo Bonzini wrote:
>> +static void edu_dma_timer(void *opaque)
>> +{
>> +EduState *edu = opaque;
>> +bool raise_irq = false;
>> +
>> +qemu_mutex_lock(&edu->dma_mutex);
>
> dma_mutex and mutex and irq_mutex are not necessary. All I/O happens
> under the big Q
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Add the bit width of every refcount entry to the format-specific
> information.
>
> In contrast to lazy_refcounts and the corrupt flag, this should be
> always emitted, even for compat=0.10 although it does not support any
> refcount width other than 16 b
Am 03.12.2014 um 13:35 hat Kevin Wolf geschrieben:
> Our IDE emulation can't handle logical block sizes other than 512. Check
> for it.
>
> The original assumption was that other values would silently be ignored
> (which is bad enough), but it's not quite true: The physical block size
> is exposed
On 12/03/2014 06:37 AM, Max Reitz wrote:
> Add two new fields regarding refcount information (the bit width of
> every entry and the maximum refcount value) to the BDRVQcowState.
>
> Signed-off-by: Max Reitz
> ---
> block/qcow2-refcount.c | 2 +-
> block/qcow2.c | 3 +++
> block/qcow2.h
Eric Blake writes:
> On 12/03/2014 03:30 AM, Kevin Wolf wrote:
>> [ CCed Benoît and Max, this is blockdev work ]
>> [ CCed Jeff, we're also talking about op blockers ]
>>
>> Not stripping quoted text for their convenience.
>
> I still intend to go through this mail in more detail, but off of a
>
On Wed, 3 Dec 2014, Don Slutz wrote:
> On 12/03/14 07:20, Stefano Stabellini wrote:
> > On Wed, 3 Dec 2014, Wei Liu wrote:
> > > On Tue, Dec 02, 2014 at 03:23:29PM -0500, Don Slutz wrote:
> > > [...]
> > > > > > > >hw_error("xc_domain_getinfo failed");
> > > > > > > >}
> > > > >
Am 03.12.2014 um 15:37 hat Markus Armbruster geschrieben:
> Kevin Wolf writes:
>
> > Our IDE emulation can't handle logical block sizes other than 512. Check
> > for it.
>
> I guess this is in part due to the 512 literals sprinkled all over the
> IDE code.
Yes, this is one part. The other part
On 12/03/2014 07:35 AM, Kevin Wolf wrote:
Our IDE emulation can't handle logical block sizes other than 512. Check
for it.
The original assumption was that other values would silently be ignored
(which is bad enough), but it's not quite true: The physical block size
is exposed in IDENTIFY DEVI
From: David Hildenbrand
We can avoid loads of syscalls when dropping to user space by storing the values
of more registers directly within kvm_run.
Support is added for:
- ARCH0: CPU timer, clock comparator, TOD programmable register,
guest breaking-event register, program parameter
- P
1 - 100 of 218 matches
Mail list logo
