From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: Juan Quintela <quintela@redhat.com>
Cc: qemu-devel@nongnu.org, amit.shah@redhat.com
Subject: Re: [Qemu-devel] [PATCH 12/17] migration: really use multiple pages at a time
Date: Fri, 3 Feb 2017 10:54:30 +0000 [thread overview]
Message-ID: <20170203105430.GB3208@work-vm> (raw)
In-Reply-To: <1485207141-1941-13-git-send-email-quintela@redhat.com>
* Juan Quintela (quintela@redhat.com) wrote:
> We now send several pages at a time each time that we wakeup a thread.
>
> Signed-off-by: Juan Quintela <quintela@redhat.com>
> ---
> migration/ram.c | 44 ++++++++++++++++++++++++++++++++++++++------
> 1 file changed, 38 insertions(+), 6 deletions(-)
>
> diff --git a/migration/ram.c b/migration/ram.c
> index 9d7bc64..1267730 100644
> --- a/migration/ram.c
> +++ b/migration/ram.c
> @@ -391,6 +391,13 @@ void migrate_compress_threads_create(void)
>
> /* Multiple fd's */
>
> +
> +typedef struct {
> + int num;
> + int size;
> + uint8_t **address;
> +} multifd_pages_t;
The naming is odd for QEMU; should be MultiFDPages ?
You might want to make num & size unsigned.
I was trying to understand why you store 'size' - is that because you worry about
someone changing the size parameter while we're running? But given that we call
init in a few places I'm not sure it covers it.
> struct MultiFDSendParams {
> /* not changed */
> QemuThread thread;
> @@ -400,7 +407,7 @@ struct MultiFDSendParams {
> /* protected by param mutex */
> bool quit;
> bool started;
> - uint8_t *address;
> + multifd_pages_t pages;
> /* protected by multifd mutex */
> bool done;
> };
> @@ -424,8 +431,8 @@ static void *multifd_send_thread(void *opaque)
>
> qemu_mutex_lock(¶ms->mutex);
> while (!params->quit){
> - if (params->address) {
> - params->address = 0;
> + if (params->pages.num) {
> + params->pages.num = 0;
> qemu_mutex_unlock(¶ms->mutex);
> qemu_mutex_lock(&multifd_send_mutex);
> params->done = true;
> @@ -473,6 +480,13 @@ void migrate_multifd_send_threads_join(void)
> multifd_send = NULL;
> }
>
> +static void multifd_init_group(multifd_pages_t *pages)
> +{
> + pages->num = 0;
> + pages->size = migrate_multifd_group();
> + pages->address = g_malloc0(pages->size * sizeof(uint8_t *));
g_new0(uint8_t *, pages->size)
> +}
> +
> void migrate_multifd_send_threads_create(void)
> {
> int i, thread_count;
> @@ -491,7 +505,7 @@ void migrate_multifd_send_threads_create(void)
> multifd_send[i].quit = false;
> multifd_send[i].started = false;
> multifd_send[i].done = true;
> - multifd_send[i].address = 0;
> + multifd_init_group(&multifd_send[i].pages);
> multifd_send[i].c = socket_send_channel_create();
> if(!multifd_send[i].c) {
> error_report("Error creating a send channel");
> @@ -511,8 +525,22 @@ void migrate_multifd_send_threads_create(void)
>
> static int multifd_send_page(uint8_t *address)
Can you comment multifd_send_page to explain what it returns.
(Do we really need u16 for fd number? More than 256 streams would seem
surprising).
> {
> - int i, thread_count;
> + int i, j, thread_count;
> bool found = false;
> + static multifd_pages_t pages;
> + static bool once = false;
> +
> + if (!once) {
> + multifd_init_group(&pages);
> + once = true;
> + }
> +
> + pages.address[pages.num] = address;
> + pages.num++;
> +
> + if (pages.num < (pages.size - 1)) {
> + return UINT16_MAX;
> + }
>
> thread_count = migrate_multifd_threads();
> qemu_mutex_lock(&multifd_send_mutex);
> @@ -530,7 +558,11 @@ static int multifd_send_page(uint8_t *address)
> }
> qemu_mutex_unlock(&multifd_send_mutex);
> qemu_mutex_lock(&multifd_send[i].mutex);
> - multifd_send[i].address = address;
> + multifd_send[i].pages.num = pages.num;
> + for(j = 0; j < pages.size; j++) {
> + multifd_send[i].pages.address[j] = pages.address[j];
> + }
> + pages.num = 0;
> qemu_cond_signal(&multifd_send[i].cond);
> qemu_mutex_unlock(&multifd_send[i].mutex);
>
> --
> 2.9.3
Dave
>
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2017-02-03 10:54 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-23 21:32 [Qemu-devel] [PATCH 00/17] multifd v3 Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 01/17] migration: transform remained DPRINTF into trace_ Juan Quintela
2017-01-24 2:20 ` Eric Blake
2017-01-24 12:20 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 02/17] migration: create Migration Incoming State at init time Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 03/17] migration: Test for disabled features on reception Juan Quintela
2017-01-24 10:33 ` Dr. David Alan Gilbert
2017-02-09 17:12 ` Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 04/17] migration: Don't create decompression threads if not enabled Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 05/17] migration: Add multifd capability Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 06/17] migration: Create x-multifd-threads parameter Juan Quintela
2017-02-02 15:06 ` Eric Blake
2017-02-09 17:28 ` Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 07/17] migration: Create x-multifd-group parameter Juan Quintela
2017-01-26 11:47 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 08/17] migration: create multifd migration threads Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 09/17] migration: Start of multiple fd work Juan Quintela
2017-01-27 17:45 ` Dr. David Alan Gilbert
2017-02-13 16:34 ` Juan Quintela
2017-02-13 16:39 ` Dr. David Alan Gilbert
2017-02-13 17:35 ` Daniel P. Berrange
2017-02-15 14:46 ` Dr. David Alan Gilbert
2017-02-15 15:01 ` Daniel P. Berrange
2017-01-23 21:32 ` [Qemu-devel] [PATCH 10/17] migration: create ram_multifd_page Juan Quintela
2017-01-27 18:02 ` Dr. David Alan Gilbert
2017-01-30 10:06 ` Juan Quintela
2017-02-02 11:04 ` Dr. David Alan Gilbert
2017-02-13 16:36 ` Juan Quintela
2017-02-14 11:26 ` Dr. David Alan Gilbert
2017-02-02 11:20 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 11/17] migration: Create thread infrastructure for multifd send side Juan Quintela
2017-01-26 12:38 ` Paolo Bonzini
2017-02-13 16:38 ` Juan Quintela
2017-02-02 12:03 ` Dr. David Alan Gilbert
2017-02-13 16:40 ` Juan Quintela
2017-02-14 11:58 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 12/17] migration: really use multiple pages at a time Juan Quintela
2017-02-03 10:54 ` Dr. David Alan Gilbert [this message]
2017-02-13 16:47 ` Juan Quintela
2017-01-23 21:32 ` [Qemu-devel] [PATCH 13/17] migration: Send the fd number which we are going to use for this page Juan Quintela
2017-02-03 10:59 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 14/17] migration: Create thread infrastructure for multifd recv side Juan Quintela
2017-01-26 12:39 ` Paolo Bonzini
2017-02-03 11:24 ` Dr. David Alan Gilbert
2017-02-13 16:56 ` Juan Quintela
2017-02-14 11:34 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 15/17] migration: Test new fd infrastructure Juan Quintela
2017-02-03 11:36 ` Dr. David Alan Gilbert
2017-02-13 16:57 ` Juan Quintela
2017-02-14 11:05 ` Dr. David Alan Gilbert
2017-02-14 11:15 ` Daniel P. Berrange
2017-01-23 21:32 ` [Qemu-devel] [PATCH 16/17] migration: [HACK]Transfer pages over new channels Juan Quintela
2017-02-03 11:41 ` Dr. David Alan Gilbert
2017-01-23 21:32 ` [Qemu-devel] [PATCH 17/17] migration: flush receive queue Juan Quintela
2017-02-03 12:28 ` Dr. David Alan Gilbert
2017-02-13 17:13 ` Juan Quintela
2017-01-23 22:12 ` [Qemu-devel] [PATCH 00/17] multifd v3 no-reply
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=20170203105430.GB3208@work-vm \
--to=dgilbert@redhat.com \
--cc=amit.shah@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.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).