On Sat, Jan 18, 2025 at 07:15:56PM +0900, Akihiko Odaki wrote:
> On 2025/01/18 2:46, Peter Xu wrote:
> > On Fri, Jan 17, 2025 at 03:24:34PM +0900, Akihiko Odaki wrote:
> > > On 2025/01/16 23:33, Peter Xu wrote:
> > > > On Thu, Jan 16, 2025 at 02:37:38PM +0900, Akihiko Odaki wrote:
> > > > > On 2025
On 2025/01/18 2:46, Peter Xu wrote:
On Fri, Jan 17, 2025 at 03:24:34PM +0900, Akihiko Odaki wrote:
On 2025/01/16 23:33, Peter Xu wrote:
On Thu, Jan 16, 2025 at 02:37:38PM +0900, Akihiko Odaki wrote:
On 2025/01/16 1:14, Peter Xu wrote:
On Thu, Jan 16, 2025 at 12:52:56AM +0900, Akihiko Odaki wr
On Fri, Jan 17, 2025 at 03:24:34PM +0900, Akihiko Odaki wrote:
> On 2025/01/16 23:33, Peter Xu wrote:
> > On Thu, Jan 16, 2025 at 02:37:38PM +0900, Akihiko Odaki wrote:
> > > On 2025/01/16 1:14, Peter Xu wrote:
> > > > On Thu, Jan 16, 2025 at 12:52:56AM +0900, Akihiko Odaki wrote:
> > > > > Functio
On 2025/01/17 1:13, BALATON Zoltan wrote:
On Thu, 16 Jan 2025, Peter Maydell wrote:
On Tue, 14 Jan 2025 at 19:12, Peter Xu wrote:
On Tue, Jan 14, 2025 at 05:42:57PM +, Peter Maydell wrote:
There's at least one test in the arm qtests that will hit this.
I suspect that you'll find that mos
On 2025/01/16 23:33, Peter Xu wrote:
On Thu, Jan 16, 2025 at 02:37:38PM +0900, Akihiko Odaki wrote:
On 2025/01/16 1:14, Peter Xu wrote:
On Thu, Jan 16, 2025 at 12:52:56AM +0900, Akihiko Odaki wrote:
Functionally, the ordering of container/subregion finalization matters if
some device tries to
On Thu, Jan 16, 2025 at 02:50:26PM +, Peter Maydell wrote:
> On Tue, 14 Jan 2025 at 19:12, Peter Xu wrote:
> >
> > On Tue, Jan 14, 2025 at 05:42:57PM +, Peter Maydell wrote:
> > > There's at least one test in the arm qtests that will hit this.
> > > I suspect that you'll find that most arc
On Thu, 16 Jan 2025, Peter Maydell wrote:
On Tue, 14 Jan 2025 at 19:12, Peter Xu wrote:
On Tue, Jan 14, 2025 at 05:42:57PM +, Peter Maydell wrote:
There's at least one test in the arm qtests that will hit this.
I suspect that you'll find that most architectures except x86
(where we don't
On Tue, 14 Jan 2025 at 19:12, Peter Xu wrote:
>
> On Tue, Jan 14, 2025 at 05:42:57PM +, Peter Maydell wrote:
> > There's at least one test in the arm qtests that will hit this.
> > I suspect that you'll find that most architectures except x86
> > (where we don't have models of complex SoCs and
On Thu, Jan 16, 2025 at 02:37:38PM +0900, Akihiko Odaki wrote:
> On 2025/01/16 1:14, Peter Xu wrote:
> > On Thu, Jan 16, 2025 at 12:52:56AM +0900, Akihiko Odaki wrote:
> > > Functionally, the ordering of container/subregion finalization matters if
> > > some device tries to a container during final
On 2025/01/16 1:14, Peter Xu wrote:
On Thu, Jan 16, 2025 at 12:52:56AM +0900, Akihiko Odaki wrote:
Functionally, the ordering of container/subregion finalization matters if
some device tries to a container during finalization. In such a case,
|
^ so
On Thu, Jan 16, 2025 at 12:52:56AM +0900, Akihiko Odaki wrote:
> Functionally, the ordering of container/subregion finalization matters if
> some device tries to a container during finalization. In such a case,
|
^ something is missing here, feel free to
On 2025/01/16 0:40, Peter Xu wrote:
On Wed, Jan 15, 2025 at 11:54:56PM +0900, Akihiko Odaki wrote:
On 2025/01/15 22:43, Peter Xu wrote:
On Wed, Jan 15, 2025 at 01:46:29PM +0900, Akihiko Odaki wrote:
On 2025/01/15 2:02, Peter Xu wrote:
On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wr
On Wed, Jan 15, 2025 at 11:54:56PM +0900, Akihiko Odaki wrote:
> On 2025/01/15 22:43, Peter Xu wrote:
> > On Wed, Jan 15, 2025 at 01:46:29PM +0900, Akihiko Odaki wrote:
> > > On 2025/01/15 2:02, Peter Xu wrote:
> > > > On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
> > > > > memory_
On 2025/01/15 22:43, Peter Xu wrote:
On Wed, Jan 15, 2025 at 01:46:29PM +0900, Akihiko Odaki wrote:
On 2025/01/15 2:02, Peter Xu wrote:
On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
memory_region_finalize() is not a function to tell the owner is leaving, but
the memory region
On Wed, Jan 15, 2025 at 01:46:29PM +0900, Akihiko Odaki wrote:
> On 2025/01/15 2:02, Peter Xu wrote:
> > On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
> > > memory_region_finalize() is not a function to tell the owner is leaving,
> > > but
> > > the memory region itself is being d
On 2025/01/15 2:02, Peter Xu wrote:
On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
memory_region_finalize() is not a function to tell the owner is leaving, but
the memory region itself is being destroyed.
It is when the lifecycle of the MR is the same as the owner. That holds
On Tue, Jan 14, 2025 at 05:42:57PM +, Peter Maydell wrote:
> On Tue, 14 Jan 2025 at 17:02, Peter Xu wrote:
> >
> > On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
> > > memory_region_finalize() is not a function to tell the owner is leaving,
> > > but
> > > the memory region it
On Tue, 14 Jan 2025 at 17:02, Peter Xu wrote:
>
> On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
> > memory_region_finalize() is not a function to tell the owner is leaving, but
> > the memory region itself is being destroyed.
>
> It is when the lifecycle of the MR is the same as t
On Tue, Jan 14, 2025 at 05:43:09PM +0900, Akihiko Odaki wrote:
> memory_region_finalize() is not a function to tell the owner is leaving, but
> the memory region itself is being destroyed.
It is when the lifecycle of the MR is the same as the owner. That holds
true I suppose if without this patch
On 2025/01/14 0:57, Peter Xu wrote:
On Sat, Jan 11, 2025 at 01:15:24PM +0900, Akihiko Odaki wrote:
On 2025/01/11 0:18, Peter Xu wrote:
On Fri, Jan 10, 2025 at 05:43:15PM +0900, Akihiko Odaki wrote:
On 2025/01/10 4:37, Peter Xu wrote:
On Thu, Jan 09, 2025 at 02:29:21PM -0500, Peter Xu wrote:
On Sat, Jan 11, 2025 at 01:15:24PM +0900, Akihiko Odaki wrote:
> On 2025/01/11 0:18, Peter Xu wrote:
> > On Fri, Jan 10, 2025 at 05:43:15PM +0900, Akihiko Odaki wrote:
> > > On 2025/01/10 4:37, Peter Xu wrote:
> > > > On Thu, Jan 09, 2025 at 02:29:21PM -0500, Peter Xu wrote:
> > > > > On Thu, Jan 0
On 2025/01/11 0:18, Peter Xu wrote:
On Fri, Jan 10, 2025 at 05:43:15PM +0900, Akihiko Odaki wrote:
On 2025/01/10 4:37, Peter Xu wrote:
On Thu, Jan 09, 2025 at 02:29:21PM -0500, Peter Xu wrote:
On Thu, Jan 09, 2025 at 01:30:35PM +0100, BALATON Zoltan wrote:
On Thu, 9 Jan 2025, Akihiko Odaki wr
On Fri, Jan 10, 2025 at 05:43:15PM +0900, Akihiko Odaki wrote:
> On 2025/01/10 4:37, Peter Xu wrote:
> > On Thu, Jan 09, 2025 at 02:29:21PM -0500, Peter Xu wrote:
> > > On Thu, Jan 09, 2025 at 01:30:35PM +0100, BALATON Zoltan wrote:
> > > > On Thu, 9 Jan 2025, Akihiko Odaki wrote:
> > > > > Do not
On 2025/01/10 4:37, Peter Xu wrote:
On Thu, Jan 09, 2025 at 02:29:21PM -0500, Peter Xu wrote:
On Thu, Jan 09, 2025 at 01:30:35PM +0100, BALATON Zoltan wrote:
On Thu, 9 Jan 2025, Akihiko Odaki wrote:
Do not refer to "memory region's reference count"
-
On Thu, Jan 09, 2025 at 02:29:21PM -0500, Peter Xu wrote:
> On Thu, Jan 09, 2025 at 01:30:35PM +0100, BALATON Zoltan wrote:
> > On Thu, 9 Jan 2025, Akihiko Odaki wrote:
> > > Do not refer to "memory region's reference count"
> > > -
> > >
> > > Now M
On Thu, Jan 09, 2025 at 01:30:35PM +0100, BALATON Zoltan wrote:
> On Thu, 9 Jan 2025, Akihiko Odaki wrote:
> > Do not refer to "memory region's reference count"
> > -
> >
> > Now MemoryRegions do have their own reference counts, but they will not
> >
On Thu, 9 Jan 2025, Akihiko Odaki wrote:
Do not refer to "memory region's reference count"
-
Now MemoryRegions do have their own reference counts, but they will not
be used when their owners are not themselves. However, the documentation
of memory_
Do not refer to "memory region's reference count"
-
Now MemoryRegions do have their own reference counts, but they will not
be used when their owners are not themselves. However, the documentation
of memory_region_ref() says it adds "1 to a memory re
28 matches
Mail list logo
