On Wed, Nov 27, 2024 at 11:12:09AM -0300, Fabiano Rosas wrote:
> Prasad Pandit writes:
>
> > On Wed, 27 Nov 2024 at 02:49, Fabiano Rosas wrote:
> >> This patch should be just the actual refactoring on top of master, with
> >> no mention to postcopy at all.
> >
> > * Okay. We'll have to ensure th
Hello Fabiano,
On Mon, 2 Dec 2024 at 19:42, Fabiano Rosas wrote:
> > ...multifd_send/_recv_sync_main();<= do the 'flush' and
> > 'sync' mean the same thing here?
>
> No, that patch is indeed inconsistent in the terminology, good point.
> Well, flush and sync are not reserved terms, we can u
Hello Fabiano,
On Thu, 28 Nov 2024 at 18:50, Fabiano Rosas wrote:
>>> We currently have a mess of conditionals to achieve the correct
>>> combination of multifd local flushes, where we sync the local
>>> (send/recv) multifd threads between themselves, and multifd remote
>>> flushes, where we put
Prasad Pandit writes:
> On Wed, 27 Nov 2024 at 19:42, Fabiano Rosas wrote:
>> From e9110360eb0efddf6945f37c518e3cc38d12b600 Mon Sep 17 00:00:00 2001
>> From: Fabiano Rosas
>> Date: Wed, 27 Nov 2024 11:03:04 -0300
>> Subject: [PATCH] migration: Rationalize multifd flushes from ram code
>>
>> We
On Wed, 27 Nov 2024 at 17:49, Fabiano Rosas wrote:
> > On Wed, 27 Nov 2024 at 02:49, Fabiano Rosas wrote:
> >> This patch should be just the actual refactoring on top of master, with
> >> no mention to postcopy at all.
...
> It doesn't need to be a single patch submission, it could be a patch at
On Wed, 27 Nov 2024 at 19:42, Fabiano Rosas wrote:
> From e9110360eb0efddf6945f37c518e3cc38d12b600 Mon Sep 17 00:00:00 2001
> From: Fabiano Rosas
> Date: Wed, 27 Nov 2024 11:03:04 -0300
> Subject: [PATCH] migration: Rationalize multifd flushes from ram code
>
> We currently have a mess of conditi
Prasad Pandit writes:
> On Wed, 27 Nov 2024 at 02:49, Fabiano Rosas wrote:
>> This patch should be just the actual refactoring on top of master, with
>> no mention to postcopy at all.
>
> * Okay. We'll have to ensure that it is merged before multifd+postcopy change.
>
>> > +if (migra
Prasad Pandit writes:
> On Wed, 27 Nov 2024 at 02:49, Fabiano Rosas wrote:
>> This patch should be just the actual refactoring on top of master, with
>> no mention to postcopy at all.
>
> * Okay. We'll have to ensure that it is merged before multifd+postcopy change.
That's ok, just put it at th
On Wed, 27 Nov 2024 at 02:49, Fabiano Rosas wrote:
> This patch should be just the actual refactoring on top of master, with
> no mention to postcopy at all.
* Okay. We'll have to ensure that it is merged before multifd+postcopy change.
> > +if (migrate_multifd() && !migration_in_pos
Prasad Pandit writes:
> From: Prasad Pandit
>
> Refactor ram_save_target_page legacy and multifd
> functions into one. Other than simplifying it,
> it frees 'migration_ops' object from usage, so it
> is expunged.
>
> When both Multifd and Postcopy modes are enabled,
> to avoid errors, the Multif
From: Prasad Pandit
Refactor ram_save_target_page legacy and multifd
functions into one. Other than simplifying it,
it frees 'migration_ops' object from usage, so it
is expunged.
When both Multifd and Postcopy modes are enabled,
to avoid errors, the Multifd threads are active until
migration rea
11 matches
Mail list logo
