On Fri, May 16, 2014 at 9:59 AM, Laurent Desnogues
wrote:
>
> just noticed the use of some too recent glib features: g_strcmp0
> and GSequence related ones.
There is a patch upstream that removes g_sequence_lookup, which was
not linking on one platform. I haven't seen any other reports of
breakag
On Wed, May 14, 2014 at 11:15 AM, Kevin Wolf wrote:
>> freeing the extra clusters.
>
> Do you have an easy reproducer? Because I can't see the bug.
Thanks for the review! I was having a hard time reproducing this until
I did a bisect. This bug was fixed by 65f33bc which was merged at or
after th
On Wed, May 14, 2014 at 9:43 AM, Jeff Cody wrote:
> Prior to a recent commit, this function did not make distinction on
> duplicates. As of commit e855e4fb7, duplicates are not longer printed
> in the help message:
>
> e855e4fb7: Ignore duplicate or NULL format_name in bdrv_iterate_format):
>
Th
On Wed, May 14, 2014 at 9:02 AM, Stefan Hajnoczi wrote:
> Yes, we cannot change the git commit history once a commit is in the
> public qemu.git repository. The only options are to:
>
> 1. Add a patch on top that fixes the issue.
> 2. Use git-revert(1) to undo your commit entirely (this adds a ne
On Wed, May 14, 2014 at 8:35 AM, Stefan Hajnoczi wrote:
> Jeff Cody recently wanted to eliminate duplicate entries in the list. I
> thought part of your intention was to address the duplicates with your
> patch.
>
> We can back out the sequence API if it's not supported on older glib but
> it wou
Signed-off-by: Mike Day
---
qemu-img.c | 7 ++-
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
index 04ce02a..1ad899e 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -70,11 +70,8 @@ static void add_format_to_seq(void *opaque, const char
*fmt_name)
{
G
ned-off-by: Mike Day
---
qemu-img.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
index 04ce02a..bf5e74c 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -70,11 +70,8 @@ static void add_format_to_seq(void *opaque, const char
*fmt_name)
{
GSequence *seq = opa
On Tue, May 13, 2014 at 10:18 AM, Cornelia Huck
wrote:
>
> qemu-img.o: In function `add_format_to_seq':
> /home/cohuck/git/qemu/qemu-img.c:73: undefined reference to
> `g_sequence_lookup'
> collect2: ld returned 1 exit status
>
> g_sequence_lookup has been added with glib 2.28, and this box has
>
ind a specific format in the list.
[Added suggestions from Fam Zheng to declare variables
at the top of the scope in help() and to omit explicit cast for void*
opaque.
--Stefan]
[Removed call to g_sequence_lookup because it breaks the build on
machines with glib < 2.28.
--Mike]
Signed-off-
moving the last snapshot there are no existing snapshots to be
shifted. All that needs to be done is to write the header and
unallocate the blocks.
Signed-off-by: Mike Day
Reviewed-by: Eric Blake
---
v2: improved the git log entry
added Eric Blake as a reviewer
block/qcow2-snapshot.
On Mon, May 12, 2014 at 12:39 PM, Eric Blake wrote:
> This information is actually quite useful in understanding the patch,
> and I would have included it prior to the --- for inclusion in git,
> rather than in the reviewer-only commentary that gets stripped during
> 'git am'.
>
>>
>> block/qcow2
On Mon, May 12, 2014 at 12:36 PM, Kevin Wolf wrote:
> What would qemu-img rebase do with -o? It is just for (safely) changing
> the backing file, not for updating options. There is qemu-img amend for
> that, and it does have an -o option.
A little background ... I'm writing a 4-day KVM training
On Mon, May 12, 2014 at 11:53 AM, Eric Blake wrote:
> Ah, but it is:
>
> $ qemu-img create -f qcow2 -o help
> Supported options:
> size Virtual disk size
> compat Compatibility level (0.10 or 1.1)
> backing_file File name of a base image
> backing_fmt Image format of
On Mon, May 12, 2014 at 11:20 AM, Eric Blake wrote:
> Online where? In 'qemu-img --help', or on some web page (at what URL)?
http://qemu.weilnetz.de/qemu-doc.html#vm_005fsnapshots
-o backing file is not documented in the help command.
Mike
For the create subcommand the backing file (-b) option is documented
on-line but not in the binary. Add it.
Signed-off-by: Mike Day
---
qemu-img-cmds.hx | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/qemu-img-cmds.hx b/qemu-img-cmds.hx
index d029609..7724709 100644
When deleting the last snapshot, copying the resulting snapshot table
currently fails, causing the delete operation to also fail. Fix the
failure by skipping the copy and just writing the snapshot header and
freeing the extra clusters.
Signed-off-by: Mike Day
---
There are two specific problems
On Thu, May 8, 2014 at 11:12 PM, Alexey Kardashevskiy wrote:
>> There are
>> a couple ways to mitigate this type of situation by using alternative
>> data structures to inform the loop traversal. I don't know if it is
>> worth the effort, though.
>
> Here I lost you :)
If I read the code correctl
On Thu, May 8, 2014 at 10:02 AM, Peter Maydell wrote:
>>> total: 0 errors, 0 warnings, 79 lines checked
>>>
>>> /tmp/a has no obvious style problems and is ready for submission.
>>> Check 500, Bad 52
>>
>> How does your new checkpatch compare to our current one?
>
> Yes; we do sometimes let checkp
On Wed, May 7, 2014 at 5:16 PM, Hani Benhabiles wrote:
> FWIW, this new version doesn't trigger a false positive I was having with
> patches 05-07 in [1].
>
> However, from a quick test for this patch on a couple of patches, I am getting
> warnings like this: "WARNING: braces {} are not necessary
Alexey Kardashevskiy writes:
> PAPR allows having multiple interrupr servers.
>
typo above in the commit log entry,
Mike
--
Mike Day | "Endurance is a Virtue"
+#define XICS_FLAGS_MSI 0x2
(nit) typo in the above comment
Mike
--
Mike Day | "Endurance is a Virtue"
On Tue, May 6, 2014 at 2:28 PM, Peter Maydell wrote:
> I think this is going to be difficult to review, to say the least.
>
> Where does your patch come from? Is the kernel's checkpatch.pl
> just a single commit between 0.31 and 0.32 (surely not) or
> a series of fixes?
Yes, this is today's versi
ind a specific format in the list.
v2: Incorporated feedback from Stefan Hajnoczi
Signed-off-by: Mike Day
---
qemu-img.c | 27 ---
1 file changed, 24 insertions(+), 3 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
index 96f4463..a559108 100644
--- a/qemu-img.c
+++ b
Thanks for the review! I've been able to shorten the next version quite a bit.
On Mon, May 5, 2014 at 9:42 AM, Stefan Hajnoczi wrote:
>> +static void GFunc_print_format(gpointer data, gpointer user)
>
> QEMU coding style is lowercase function and variable names. The
> scripts/checkpatch.pl scri
ind a specific format in the list.
Signed-off-by: Mike Day
---
qemu-img.c | 33 -
1 file changed, 32 insertions(+), 1 deletion(-)
diff --git a/qemu-img.c b/qemu-img.c
index 96f4463..d8b7ef4 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -32,6 +32,7 @@
#in
On Wed, Apr 30, 2014 at 10:16 AM, Eric Blake wrote:
> Additional patches:
> https://lists.gnu.org/archive/html/qemu-devel/2014-04/msg04074.html
> https://lists.gnu.org/archive/html/qemu-devel/2014-04/msg04468.html
> both already on the queue for the next PULL request of the block branch.
Excellen
On Sun, Apr 27, 2014 at 9:55 PM, Fam Zheng wrote:
>> > /* not found */
>> > -help();
>> > -return 0;
>> > +error_exit("Command not found: %s", cmdname);
>>
>> Looks like we just relied previously on the default 'not found' case
>> for help() to provide the "--help" option.
>>
>
>
On Tue, Apr 29, 2014 at 1:02 AM, Michael Tokarev wrote:
> I guess we should have some global variable like "stdio_occuped", set
> it to 1 when -daemonize is specified, and set and check it each time
> we try to use stdio for something. This way we'll prevent various
> parts of qemu from fighting
On Tue, Apr 29, 2014 at 3:09 AM, Gerd Hoffmann wrote:
>
> I don't feel like adding more band-aid to paper over the fundamental
> issue. Too much band-aid tends to cause other unwanted side effects.
Fair enough, I tend to agree. thanks for the review.
Mike
On Mon, Apr 28, 2014 at 3:53 PM, Michael Tokarev wrote:
> Gosh. This is uuugly.
>
> Maybe, at least, we can add some variable which gets incremented for
> each -mon chardev= ? I dunno, that's about the same ugliness. Oh
> well... :(
Suggestions much appreciated, I think its ugly versus intr
checking for the existing chardev before adding another.
Signed-off-by: Mike Day
---
To reproduce, use -mon and -nographic together. I was able to
reproduce it using
# qemu-system-x86_64 -enable-kvm -m 1G -chardev stdio,id=mon0 \
# -mon chardev=mon0,mode=readline -nographic
---
vl.c | 22
On Thu, Apr 24, 2014 at 3:31 AM, Markus Armbruster wrote:
>> I believe someone on the list mentioned they are seeing a couple
>> problems entering and exiting the Monitor. I'd like to look at this more
>> closely, starting with my most pending issue: losing the terminal echo
>> after exiting the M
look for this
code? Otherwise I'll start looking through main_loop and friends and
vl.c for init and destroy routines.
Mike
--
Mike Day | "Endurance is a Virtue"
On Tue, Apr 8, 2014 at 2:28 AM, Alexander Graf wrote:
> On 22.03.14 14:03, Alexey Kardashevskiy wrote:
>>
>> This adds handling of the RESOURCE_ADDR_TRANS_MODE resource from
>> the H_SET_MODE, for POWER8 (PowerISA 2.07) only.
>>
>> Signed-off-by: Alexey Kardas
Alexey Kardashevskiy writes:
> This adds handling of the RESOURCE_ADDR_TRANS_MODE resource from
> the H_SET_MODE, for POWER8 (PowerISA 2.07) only.
>
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> hw/ppc/spapr_hcall.c | 26 ++
&
void rtas_st_buffer(target_ulong phys,
> + target_ulong phys_len,
> + uint8_t *buffer) {
> +uint32_t i;
> +for (i = 0; i < phys_len; i++) {
> +stb_phys(ppc64_phys_to_real(phys + i), buffer[i]);
> +
edef void (*spapr_rtas_fn)(PowerPCCPU *cpu, sPAPREnvironment *spapr,
>uint32_t token,
>uint32_t nargs, target_ulong args,
> diff --git a/pixman b/pixman
> --- a/pixman
> +++ b/pixman
> @@ -1 +1 @@
> -Subproject commit 97336fad32acf802003855cd8bd6477fa49a12e3
> +Subproject commit 97336fad32acf802003855cd8bd6477fa49a12e3-dirty
--
Mike Day | "Endurance is a Virtue"
his reworks the H_PUT_TCE handler to make use of the new put_tce()
> callback.
>
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> hw/ppc/spapr_iommu.c | 21 +
> include/hw/ppc/spapr.h | 13 +
> 2 files changed, 30 insertions(
On 21/11/13 15:08 +1100, Alexey Kardashevskiy wrote:
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> hw/ppc/spapr_pci.c | 8
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/hw/ppc/spapr_pci.c b/hw/ppc/spapr_pci.c
> index 96
h_realize()
> call at the end of the spapr_finalize().
>
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> Changes:
> v5:
> * this is a new patch in the series, it was a part of a previous patch
> ---
> hw/ppc/spapr_pci.c | 46
> +++
lize() does not "return" any value, instead it puts the extended
> error into **errp.
>
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> Changes:
> v5:
> * finish_finalize() moved to a separate patch
> ---
> hw/ppc/spapr_pci.c | 44 ++
lize() does not "return" any value, instead it puts the extended
> error into **errp.
>
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> Changes:
> v5:
> * finish_finalize() moved to a separate patch
> ---
> hw/ppc/spapr_pci.c | 44 ++
se
RCU for read operations V3."
V3:
- timerlists modifications split to a separate patch (this one).
- Addressed comments from Alex Bligh and Paolo Bonzini.
Signed-off-by: Mike Day
---
qemu-timer.c | 31 ++-
1 file changed, 26 insertions(+), 5 deletions(-)
di
ff-by: Mike Day
---
include/qemu/timer.h | 19 +
qemu-timer.c | 111 +--
2 files changed, 72 insertions(+), 58 deletions(-)
diff --git a/include/qemu/timer.h b/include/qemu/timer.h
index 5afcffc..f69ec49 100644
--- a/include/qemu/ti
The first patch in the series converts the
clock->timerlists->active_timers list to RCU. It also creates
QemuMutex in the parent QemuClock data structure.
The second patch converts clock->timerlists to RCU.
Mike Day (2):
[RFC] Convert active timers list to use RCU V3
[RFC] Conv
Alexey Kardashevskiy writes:
> This adds migration support for OHCI.
>
> Signed-off-by: Alexey Kardashevskiy
Reviewed-by: Mike Day
> ---
> hw/usb/hcd-ohci.c | 12
> 1 file changed, 12 insertions(+)
>
> diff --git a/hw/usb/hcd-ohci.c b/hw/usb/hcd-ohci.c
s @cpu_index fixup as @cpu_dt_id is used instead so QEMU monitor
> can accept command-line CPU indexes again.
>
> This changes kvm_arch_vcpu_id() to use ppc_get_vcpu_dt_id() as at the moment
> KVM CPU id and device tree ID are calculated using the same algorithm.
>
Signed-off-by: Ale
alized from @cpu_index by default and can be fixed later
> to meet the device tree requirements.
>
> This adds an API to handle @cpu_dt_id.
>
> This removes kvmppc_fixup_cpu() as it is not more needed, @cpu_dt_id
> is calculated in ppc_cpu_realize().
>
> This will be used
On Fri, Feb 28, 2014 at 3:05 PM, Alex Bligh wrote:
>> Rather than introduce a second mutex for timerlists, which would
>> require nested mutexes to in orderwrite to the timerlists, use one
>> QemuMutex in the QemuClock structure for all write access to any list
>> hanging off the QemuClock struct
anch:
https://github.com/bonzini/qemu/tree/rcu and also requires the
previously submitted patch 03fba95 "Convert active timers list to use
RCU for read operations V2."
V2:
- timerlists modifications split to a separate patch (this one).
- Addressed Alex Blighs comments.
Signed-off-by: Mike Day
-
The first patch in the series convers the
clock->timerlists->active_timers list to RCU. The second patch
converts clock->timerlists to RCU and also protects access to
timerlists->active_timers->timer_list.
Mike Day (2):
[RFC] Convert active timers list to use RCU V2
[RFC
b.com/bonzini/qemu/tree/rcu
V2:
- Addresses comments from Alex Bligh
Signed-off-by: Mike Day
---
include/qemu/timer.h | 19 +++
qemu-timer.c | 69 ++--
2 files changed, 44 insertions(+), 44 deletions(-)
diff --git a/include/qemu/ti
> 1. You seem to be removing the use of the active_timers_lock and replacing it
> by
>rcu (fine). However, you seem to have left the qemu_mutex_destroy in
>timerlist_free, and left the mutex in QEMUTimerList. Any reason why we
> need both?
>
I responded incorrectly to this yesterday. We
but you appear to have
>introduced a rcu_read_lock() outside the loop, but then
>unlock it within the loop. If the loop iterates more than once,
>won't it do more unlocks than locks?
Yes, I think the right fix is to omit the last rcu_read_unlock.
--
Mike Day | "Endurance is a Virtue"
On Thu, Feb 13, 2014 at 4:25 AM, Paolo Bonzini wrote:
> Il 13/02/2014 10:11, Alex Bligh ha scritto:
>>
>> I'll certainly have a look through this. However before I do, what
>> problem is this trying to solve? Do we think there is possibility
>> of contention on the active timers lock? I used to th
Allow readers to use RCU when reading Qemu timer lists. Applies to
Paolo Bonzini's RCU branch, https://github.com/bonzini/qemu/tree/rcu.
This patch is for comment and review only. The rcu branch needs to be
rebased on upstream.
Signed-off-by: Mike Day
---
include/qemu/timer.h
fig space back to zero
> 3) EXIT:exit() from qemu with a status code
> +static uint32_t virtio_testdev_get_features(VirtIODevice *vdev, uint32_t f)
> +{
> +return f;
> +}
> +
Is this meant to be a stub currently?
Mike
--
Mike Day | "Endurance is a Virtue"
On Mon, Jan 20, 2014 at 12:24 PM, Michael Roth
wrote:
> Quoting Alexey Kardashevskiy (2014-01-19 20:58:20)
>
> Would need to look at it a bit more closely to say for certain, but after
> discussing it a bit Tyrel/Mike, I think the main considerations would be:
>
> 1) PHB hotplug/unplug would need
Do you know which device is writing to the BAR below? From the trace
it appears it should be restoring the memory address to the BAR after
writing all 1s to the BAR and reading back the contents. (the protocol
for finding the length of the bar memory.)
On Thu, Jan 9, 2014 at 12:24 PM, Alex William
On Sun, Jan 19, 2014 at 9:58 PM, Alexey Kardashevskiy wrote:
>
> I did not realize DRC is not just for PCI. How hard would it be to add hot
> plug support for a whole PHB? The current QEMU trend is to make QEMU
> monitor's "device_add" equal to the command line's "-device" which is not
> (yet) tru
>>
>> The address above has already been masked. What you need to do is read
>> the BAR. If the value from the BAR end in '1', its MMIO. If it ends in
>> '10', its RAM. If it ends in '0n' its disabled. The first thing that
>> the PCI software does after reading the BAR is mask off the two low
>> bi
On Tue, Jan 14, 2014 at 12:49 PM, Mike Day wrote:
>>> > > >>>>>>>
>>> > > >>>>>>>>> Prior to this change, there was no re-map with the
>>> > > >>>>>>>>> febe
>
>> > > >>>
>> > > > Prior to this change, there was no re-map with the
>> > > > febe
> If we choose not to map them, how do we distinguish them from guest RAM?
> There's no MemoryRegion flag that I'm aware of to distinguish a ram_ptr
> that points to a chunk of gue
On Tue, Jan 14, 2014 at 9:05 AM, Michael S. Tsirkin wrote:
> On Tue, Jan 14, 2014 at 08:50:54AM -0500, Mike Day wrote:
>>
>> Also, both 32-bit and 64-bit BARs are required to be supported. It is
>> legal to construct a 64-bit BAR by masking all the high bits to
>> zero
and 64-bit BARs are required to be supported. It is
legal to construct a 64-bit BAR by masking all the high bits to
zero. Presumably it would be OK to mask the 16 high bits to zero as
well, constructing a 48-bit address.
Mike
--
Mike Day | "Endurance is a Virtue"
On Fri, Jan 10, 2014 at 9:25 AM, Alexander Graf wrote:
>>> On 01/11/2014 01:00 AM, Alexander Graf wrote:
>>
Can't we determine the number of "default threads" at a common place,
preferably derived from cpu type?
>>>
>>> We can do anything. I asked how exactly as I really (really) do not
On Fri, Jan 10, 2014 at 9:13 AM, Alexey Kardashevskiy wrote:
> On 01/11/2014 01:00 AM, Alexander Graf wrote:
>> Can't we determine the number of "default threads" at a common place,
>> preferably derived from cpu type?
>
> We can do anything. I asked how exactly as I really (really) do not
> unde
ax" is a more user-friendly option. In
my understanding this is strictly a usability issue.
Mike
--
Mike Day | "Endurance is a Virtue"
3. machine_init - creates CPUs which may or may be not "host".
The patch as it its now is very simple and well-contained. I wonder how
much it would expand if we added a max thread count to the cpu class. It
seems like the need for a max thread count is idiomatic to powerpc.
Mike
--
Mike Day | "Endurance is a Virtue"
threads = threads > 0 ? threads : 1;
> +}
> if (cpus == 0) {
> cpus = cores * threads * sockets;
> }
--
Mike Day | "Endurance is a Virtue"
69 62 36 64 6a 01 06 03 gb5dh...ib6dj...
0020: 6b 62 37 64 6c 01 07 03 6d 62 38 64 6e 01 08 03 kb7dl...mb8dn...
0030: 6f 62 39 64 70 01 09 03 71 62 78 64ob9dp...qbxd
Signed-off-by: Mike Day
---
include/qemu-common.h | 2 ++
util/hexdump.c| 48
Adding Anthony's corrected address.
On Thu, Oct 10, 2013 at 6:25 PM, Mike Day wrote:
[RFC] SPAPR-PCI Hotplug Support in Qemu
>
> Background:
> ppc64 has a unique bus structure for PCI slots: each slot is connected
> to its PHB by a pci switch. This is true in some IBM ha
phb_add_pci_dt
https://github.com/mdroth/qemu/blob/spapr-pci-hotplug/hw/ppc/spapr_pci.c#L900
[4] dlpar_pci_add_bus
http://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/drivers/pci/hotplug/rpadlpar_core.c?id=8bf3379a74bc9132751bfa685bad2da318fd59d7#n165
--
Mike Day | + 1 919
a
>> sensible option.
>
> I agree. Keeping the write side on the BQL is sane, but RCU-protecting
> the read side actually makes the rules simpler.
>
> Mike, would you like to give it a shot?
Yes, I will. I'll have a patchset for review within a couple of days.
Mike
On Mon, Sep 30, 2013 at 9:34 AM, Alex Bligh wrote:
>>> > void qemu_clock_notify(QEMUClockType type)
>
> ...
>>>
>>> > int64_t qemu_clock_deadline_ns_all(QEMUClockType type)
>
> ...
>
>>> > I think these functions are always called now with the BQL held, so I
>>> > wonder if they are good candidate
>
>
> On Mon, Sep 30, 2013 at 8:55 AM, Alex Bligh wrote:
> >
> >
> > On 30 Sep 2013, at 13:45, Mike Day wrote:
> >
> > > I've applied this set to Paolo's rcu tree - I see a couple of routines
> > > that appear to need the active_
}
> }
> -return false;
> +qemu_mutex_unlock(&timer_list->active_timers_lock);
> +return found;
> }
>
> bool timer_expired(QEMUTimer *timer_head, int64_t current_time)
> @@ -369,23 +410,31 @@ bool timerlist_run_timers(QEMUTimerList *timer_list)
>
Paolo Bonzini writes:
> Il 25/09/2013 08:27, liu ping fan ha scritto:
>> Hi, is hpet orphan? Or who can help me to merge this patch-set if my
>> patch is fine.
>
> Anthony, Michael?
Sorry, wrong Michael -
Mike
--
Mike Day | + 1 919 371-8786 | ncm...@ncultra.org
"Endurance is a Virtue"
Paolo Bonzini writes:
> Il 25/09/2013 08:27, liu ping fan ha scritto:
>> Hi, is hpet orphan? Or who can help me to merge this patch-set if my
>> patch is fine.
>
> Anthony, Michael?
Yes, happy to help out with this. I'll start looking at it now and work
with Liu P
On Mon, Sep 9, 2013 at 12:21 PM, Paolo Bonzini wrote:
>
> > @@ -601,12 +608,22 @@ static void reset_ram_globals(void)
> > last_seen_block = NULL;
> > last_sent_block = NULL;
> > last_offset = 0;
> > -last_version = ram_list.version;
> > ram_bulk_stage = true;
> > +smp_w
ing KVM -t qemu and
passes 15/15 migration tests.
To apply this patch, you must base upon Paolo Bonzini's rcu tree and
also apply the RCU DQ patch (below).
https://github.com/bonzini/qemu/tree/rcu
http://article.gmane.org/gmane.comp.emulators.qemu/230159/
Signed-off-by: Mike Day
Signed-o
sses 15/15 migration tests.
To apply this patch, you must base upon Paolo Bonzini's rcu tree and
also apply the RCU DQ patch (below).
https://github.com/bonzini/qemu/tree/rcu
http://article.gmane.org/gmane.comp.emulators.qemu/230159/
Signed-off-by: Mike Day
---
arch_init.c
On Wed, Sep 4, 2013 at 3:58 PM, Paolo Bonzini wrote:
>
> > @@ -1323,23 +1325,21 @@ static RAMBlock *qemu_get_ram_block(ram_addr_t
addr)
> > {
> > RAMBlock *block;
> >
> > -/* The list is protected by the iothread lock here. */
> > +/* This assumes the iothread lock is taken here too
u/tree/rcu
http://article.gmane.org/gmane.comp.emulators.qemu/230159/
Signed-off-by: Mike Day
---
arch_init.c | 96 +---
exec.c | 162 +--
include/exec/cpu-all.h | 13 ++--
include/qemu/rcu_queue.h | 8 +++
upon Paolo Bonzini's rcu tree and
also apply the RCU DQ patch (below).
https://github.com/bonzini/qemu/tree/rcu
http://article.gmane.org/gmane.comp.emulators.qemu/230159/
Signed-off-by: Mike Day
---
arch_init.c | 103 ++
exec.c
On Tue, Sep 3, 2013 at 10:09 AM, Paolo Bonzini wrote:
>
> Il 03/09/2013 15:56, Mike Day ha scritto:
> >> > +/* this implements a long-running RCU critical section.
> >> > + * When rcu reclaims in the code start to become numerous
> >> >
On Fri, Aug 30, 2013 at 12:38 PM, Paolo Bonzini wrote:
>
> > @@ -867,7 +879,12 @@ static int ram_load(QEMUFile *f, void *opaque, int
version_id)
> > if (version_id < 4 || version_id > 4) {
> > return -EINVAL;
> > }
> > -
> > +/* this implements a long-running RCU critical se
at at
this point.
To apply this patch, you must base upon Paolo Bonzini's rcu tree and
also apply the RCU DQ patch (below).
https://github.com/bonzini/qemu/tree/rcu
http://article.gmane.org/gmane.comp.emulators.qemu/230159/
Signed-off-by: Mike Day
---
arch_init.c
On Fri, Aug 30, 2013 at 4:19 AM, Paolo Bonzini wrote:
>
>
> I'm not sure about that; returning an RCU-protected variable after
> rcu_read_unlock() seems wrong to me because the pointer may not be valid
> at that point. I suggest using a comment that asks to call
> host_from_stream_offset within r
On Wed, Aug 28, 2013 at 12:35 PM, Paolo Bonzini wrote:
>
> > @@ -828,13 +829,18 @@ static inline void
*host_from_stream_offset(QEMUFile *f,
> > qemu_get_buffer(f, (uint8_t *)id, len);
> > id[len] = 0;
> >
> > -QTAILQ_FOREACH(block, &ram_list.blocks, next) {
> > -if (!strncmp(
x, and arm targets. It has not been tested further than that at
this point.
To apply this patch, you must base upon Paolo Bonzini's rcu tree and
also apply the RCU DQ patch (below).
https://github.com/bonzini/qemu/tree/rcu
http://article.gmane.org/gmane.comp.emulators.qemu/230159/
Signed-o
ith a single updater.
This patchset builds on top of Paolo Bonzini's rcu tree:
https://github.com/bonzini/qemu/tree/rcu
Signed-off-by: Mike Day
---
docs/rcu.txt | 2 +-
include/qemu/queue.h | 11 --
include/qemu/rcu_queue.h | 137 +++
tests/Makefile
#x27;m working with
it right now. I like the idea of using a flag to provide a form of
atomicity for the doubly-linked list elements. I'm also planning on running
some timing tests to see of the additional memory barriers and atomic
accesses make *any* difference whatsoever.
Mike
Mike Day | ncm...@ncultra.org | +1 919 371-8786
of the previous list element) this is assiging the
*previous* element's next pointer, the way I read it.
The Linux list_del_rcu is dealing with a singly linked list and
therefore does not set a value in the previous node's element.
But I'm still unclear on whether or not the memory barrier is needed
because the deleted element won't be reclaimed right away.
Mike
--
Mike Day | + 1 919 371-8786 | ncm...@ncultra.org
"Endurance is a Virtue"
to contend with a single updater.
This patchset builds on top of Paolo Bonzini's rcu tree:
https://github.com/bonzini/qemu/tree/rcu
Signed-off-by: Mike Day
---
docs/rcu.txt | 2 +-
include/qemu/queue.h | 11 --
include/qemu/rcu_queue.h | 145
interface.
To accompany the RCU-enabled DQ, there is also a test file that uses
concurrent readers to contend with a single updater.
This patchset builds on top of Paolo Bonzini's rcu tree:
https://github.com/bonzini/qemu/tree/rcu
Signed-off-by: Mike Day
---
docs/rcu.txt
Peter Maydell writes:
> On 19 August 2013 19:33, Mike Day wrote:
>> These patches apply to Paolo Bonzini's rcu tree:
>>
>> https://github.com/bonzini/qemu/tree/rcu
>> commit 781e47bf1693a80b84eec298a6a1c7b29ab2c135
>>
>> Signed-off-by: Mike Day
&
These patches change from using Linux kernel style upcasts to typesafe
object oriented casts with runtime checking semantics.
These patches apply to Paolo Bonzini's rcu tree:
https://github.com/bonzini/qemu/tree/rcu
commit 781e47bf1693a80b84eec298a6a1c7b29ab2c135
Signed-off-by: Mik
These patches apply to Paolo Bonzini's rcu tree:
https://github.com/bonzini/qemu/tree/rcu
commit 781e47bf1693a80b84eec298a6a1c7b29ab2c135
Signed-off-by: Mike Day
---
hw/misc/ivshmem.c | 2 +-
hw/pci-bridge/pci_bridge_dev.c | 6 +++---
hw/pci/pci_bridge.c| 2 +-
3
Andreas Färber writes:
>> https://github.com/ncultra/qemu/tree/rcu-for-1.7
>>
>> Mike Day (2):
>> fixup changes from commit f63ca950
>> fixup changes from commit f62a6b2f from Paulo Bonzini
>
> These two patches are certainly not acceptable for upstrea
1 - 100 of 123 matches
Mail list logo