From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: Wei Yang <richardw.yang@linux.intel.com>
Cc: qemu-devel@nongnu.org, quintela@redhat.com
Subject: Re: [PATCH 5/6] migration/postcopy: enable random order target page arrival
Date: Thu, 7 Nov 2019 09:14:05 +0000 [thread overview]
Message-ID: <20191107091405.GA2816@work-vm> (raw)
In-Reply-To: <20191107060010.GA13394@richard>
* Wei Yang (richardw.yang@linux.intel.com) wrote:
> On Wed, Nov 06, 2019 at 08:08:28PM +0000, Dr. David Alan Gilbert wrote:
> >* Wei Yang (richardw.yang@linux.intel.com) wrote:
> >> After using number of target page received to track one host page, we
> >> could have the capability to handle random order target page arrival in
> >> one host page.
> >>
> >> This is a preparation for enabling compress during postcopy.
> >>
> >> Signed-off-by: Wei Yang <richardw.yang@linux.intel.com>
> >> ---
> >> migration/ram.c | 16 +++-------------
> >> 1 file changed, 3 insertions(+), 13 deletions(-)
> >>
> >> diff --git a/migration/ram.c b/migration/ram.c
> >> index b5759793a9..da0596411c 100644
> >> --- a/migration/ram.c
> >> +++ b/migration/ram.c
> >> @@ -4015,7 +4015,6 @@ static int ram_load_postcopy(QEMUFile *f)
> >> MigrationIncomingState *mis = migration_incoming_get_current();
> >> /* Temporary page that is later 'placed' */
> >> void *postcopy_host_page = mis->postcopy_tmp_page;
> >> - void *last_host = NULL;
> >> bool all_zero = false;
> >> int target_pages = 0;
> >>
> >> @@ -4062,24 +4061,15 @@ static int ram_load_postcopy(QEMUFile *f)
> >> * that's moved into place later.
> >> * The migration protocol uses, possibly smaller, target-pages
> >> * however the source ensures it always sends all the components
> >> - * of a host page in order.
> >> + * of a host page in one chunk.
> >> */
> >> page_buffer = postcopy_host_page +
> >> ((uintptr_t)host & (block->page_size - 1));
> >> /* If all TP are zero then we can optimise the place */
> >> if (target_pages == 1) {
> >> all_zero = true;
> >> - } else {
> >> - /* not the 1st TP within the HP */
> >> - if (host != (last_host + TARGET_PAGE_SIZE)) {
> >> - error_report("Non-sequential target page %p/%p",
> >> - host, last_host);
> >> - ret = -EINVAL;
> >> - break;
> >> - }
> >
> >I think this is losing more protection than needed.
> >I think you can still protect against a page from a different host-page
> >arriving until we've placed the current host-page.
> >So something like:
> >
> > if (((uintptr_t)host & ~(block->page_size - 1)) !=
> > last_host)
> >
>
> OK, looks reasonable.
>
> >and then set last_host to the start of the host page.
> >
>
> I think it is not necessary to update the last_host on each target page. We
> can just set it at the first target page.
Yes, that would be fine.
Dave
> >Then you'll check if that flush is really working.
> >
> >Dave
> >
> >> }
> >>
> >> -
> >> /*
> >> * If it's the last part of a host page then we place the host
> >> * page
> >> @@ -4090,7 +4080,6 @@ static int ram_load_postcopy(QEMUFile *f)
> >> }
> >> place_source = postcopy_host_page;
> >> }
> >> - last_host = host;
> >>
> >> switch (flags & ~RAM_SAVE_FLAG_CONTINUE) {
> >> case RAM_SAVE_FLAG_ZERO:
> >> @@ -4143,7 +4132,8 @@ static int ram_load_postcopy(QEMUFile *f)
> >>
> >> if (!ret && place_needed) {
> >> /* This gets called at the last target page in the host page */
> >> - void *place_dest = host + TARGET_PAGE_SIZE - block->page_size;
> >> + void *place_dest = (void *)QEMU_ALIGN_DOWN((unsigned long)host,
> >> + block->page_size);
> >>
> >> if (all_zero) {
> >> ret = postcopy_place_page_zero(mis, place_dest,
> >> --
> >> 2.17.1
> >>
> >--
> >Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
>
> --
> Wei Yang
> Help you, Help me
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2019-11-07 9:16 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-18 0:48 [PATCH 0/6] migration/postcopy: enable compress during postcopy Wei Yang
2019-10-18 0:48 ` [PATCH 1/6] migration/postcopy: reduce memset when it is zero page and matches_target_page_size Wei Yang
2019-11-06 18:18 ` Dr. David Alan Gilbert
2019-10-18 0:48 ` [PATCH 2/6] migration/postcopy: wait for decompress thread in precopy Wei Yang
2019-11-06 19:57 ` Dr. David Alan Gilbert
2019-10-18 0:48 ` [PATCH 3/6] migration/postcopy: count target page number to decide the place_needed Wei Yang
2019-11-06 19:59 ` Dr. David Alan Gilbert
2019-10-18 0:48 ` [PATCH 4/6] migration/postcopy: set all_zero to true on the first target page Wei Yang
2019-11-06 20:04 ` Dr. David Alan Gilbert
2019-10-18 0:48 ` [PATCH 5/6] migration/postcopy: enable random order target page arrival Wei Yang
2019-11-06 20:08 ` Dr. David Alan Gilbert
2019-11-07 6:00 ` Wei Yang
2019-11-07 9:14 ` Dr. David Alan Gilbert [this message]
2019-10-18 0:48 ` [PATCH 6/6] migration/postcopy: enable compress during postcopy Wei Yang
2019-11-06 19:55 ` Dr. David Alan Gilbert
2019-10-18 16:50 ` [PATCH 0/6] " no-reply
2019-10-19 0:15 ` Wei Yang
2019-11-06 20:11 ` Dr. David Alan Gilbert
2019-11-07 6:02 ` Wei Yang
2019-11-07 9:15 ` Dr. David Alan Gilbert
2019-11-07 12:03 ` Wei Yang
2019-11-07 12:06 ` Dr. David Alan Gilbert
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191107091405.GA2816@work-vm \
--to=dgilbert@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=richardw.yang@linux.intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).