From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:47527) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1evQof-0003PI-Ml for qemu-devel@nongnu.org; Mon, 12 Mar 2018 13:01:00 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1evQoc-0002TP-DQ for qemu-devel@nongnu.org; Mon, 12 Mar 2018 13:00:53 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:37616 helo=mx1.redhat.com) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1evQoc-0002Sl-8v for qemu-devel@nongnu.org; Mon, 12 Mar 2018 13:00:50 -0400 Date: Mon, 12 Mar 2018 17:00:31 +0000 From: "Dr. David Alan Gilbert" Message-ID: <20180312170030.GP3219@work-vm> References: <20180308195811.24894-1-dgilbert@redhat.com> <20180308195811.24894-18-dgilbert@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH v4 17/29] vhost+postcopy: Send requests to source for shared pages List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?iso-8859-1?Q?Marc-Andr=E9?= Lureau Cc: QEMU , "Michael S. Tsirkin" , Maxime Coquelin , Peter Xu , Juan Quintela , Andrea Arcangeli * Marc-Andr=E9 Lureau (marcandre.lureau@gmail.com) wrote: > Hi >=20 > On Thu, Mar 8, 2018 at 8:57 PM, Dr. David Alan Gilbert (git) > wrote: > > From: "Dr. David Alan Gilbert" > > > > Send requests back to the source for shared page requests. > > >=20 > Well, you could update the commit message to say that it introduces a > callback to be used in the following patches, as well as saving RB of > last request. >=20 > For now that patch does no functional change :) >=20 > > Signed-off-by: Dr. David Alan Gilbert >=20 > Reviewed-by: Marc-Andr=E9 Lureau Thanks, Updated text to: vhost+postcopy: Helper to send requests to source for shared pages =20 Provide a helper to be used by shared waker functions to request shared pages from the source. The last_rb pointer is moved into the incoming state since this helper can update it as well as the main fault thread function. Dave >=20 > > --- > > migration/migration.h | 2 ++ > > migration/postcopy-ram.c | 32 +++++++++++++++++++++++++++++--- > > migration/postcopy-ram.h | 3 +++ > > migration/trace-events | 2 ++ > > 4 files changed, 36 insertions(+), 3 deletions(-) > > > > diff --git a/migration/migration.h b/migration/migration.h > > index d02a759331..83dc36b57a 100644 > > --- a/migration/migration.h > > +++ b/migration/migration.h > > @@ -49,6 +49,8 @@ struct MigrationIncomingState { > > int userfault_event_fd; > > QEMUFile *to_src_file; > > QemuMutex rp_mutex; /* We send replies from multiple threads = */ > > + /* RAMBlock of last request sent to source */ > > + RAMBlock *last_rb; > > void *postcopy_tmp_page; > > void *postcopy_tmp_zero_page; > > /* PostCopyFD's for external userfaultfds & handlers of shared m= emory */ > > diff --git a/migration/postcopy-ram.c b/migration/postcopy-ram.c > > index 4ab1b7d36d..83780576ca 100644 > > --- a/migration/postcopy-ram.c > > +++ b/migration/postcopy-ram.c > > @@ -525,6 +525,32 @@ static int ram_block_enable_notify(const char *b= lock_name, void *host_addr, > > return 0; > > } > > > > +/* > > + * Callback from shared fault handlers to ask for a page, > > + * the page must be specified by a RAMBlock and an offset in that rb > > + * Note: Only for use by shared fault handlers (in fault thread) > > + */ > > +int postcopy_request_shared_page(struct PostCopyFD *pcfd, RAMBlock *= rb, > > + uint64_t client_addr, uint64_t rb_o= ffset) > > +{ > > + size_t pagesize =3D qemu_ram_pagesize(rb); > > + uint64_t aligned_rbo =3D rb_offset & ~(pagesize - 1); > > + MigrationIncomingState *mis =3D migration_incoming_get_current()= ; > > + > > + trace_postcopy_request_shared_page(pcfd->idstr, qemu_ram_get_ids= tr(rb), > > + rb_offset); > > + /* TODO: Check bitmap to see if we already have the page */ > > + if (rb !=3D mis->last_rb) { > > + mis->last_rb =3D rb; > > + migrate_send_rp_req_pages(mis, qemu_ram_get_idstr(rb), > > + aligned_rbo, pagesize); > > + } else { > > + /* Save some space */ > > + migrate_send_rp_req_pages(mis, NULL, aligned_rbo, pagesize); > > + } > > + return 0; > > +} > > + > > /* > > * Handle faults detected by the USERFAULT markings > > */ > > @@ -535,9 +561,9 @@ static void *postcopy_ram_fault_thread(void *opaq= ue) > > int ret; > > size_t index; > > RAMBlock *rb =3D NULL; > > - RAMBlock *last_rb =3D NULL; /* last RAMBlock we sent part of */ > > > > trace_postcopy_ram_fault_thread_entry(); > > + mis->last_rb =3D NULL; /* last RAMBlock we sent part of */ > > qemu_sem_post(&mis->fault_thread_sem); > > > > struct pollfd *pfd; > > @@ -636,8 +662,8 @@ static void *postcopy_ram_fault_thread(void *opaq= ue) > > * Send the request to the source - we want to request o= ne > > * of our host page sizes (which is >=3D TPS) > > */ > > - if (rb !=3D last_rb) { > > - last_rb =3D rb; > > + if (rb !=3D mis->last_rb) { > > + mis->last_rb =3D rb; > > migrate_send_rp_req_pages(mis, qemu_ram_get_idstr(rb= ), > > rb_offset, qemu_ram_pagesiz= e(rb)); > > } else { > > diff --git a/migration/postcopy-ram.h b/migration/postcopy-ram.h > > index c8ced3470b..d7afab07d7 100644 > > --- a/migration/postcopy-ram.h > > +++ b/migration/postcopy-ram.h > > @@ -164,5 +164,8 @@ struct PostCopyFD { > > */ > > void postcopy_register_shared_ufd(struct PostCopyFD *pcfd); > > void postcopy_unregister_shared_ufd(struct PostCopyFD *pcfd); > > +/* Callback from shared fault handlers to ask for a page */ > > +int postcopy_request_shared_page(struct PostCopyFD *pcfd, RAMBlock *= rb, > > + uint64_t client_addr, uint64_t offs= et); > > > > #endif > > diff --git a/migration/trace-events b/migration/trace-events > > index 1e617ad7a6..7c910b5479 100644 > > --- a/migration/trace-events > > +++ b/migration/trace-events > > @@ -198,6 +198,8 @@ postcopy_ram_incoming_cleanup_closeuf(void) "" > > postcopy_ram_incoming_cleanup_entry(void) "" > > postcopy_ram_incoming_cleanup_exit(void) "" > > postcopy_ram_incoming_cleanup_join(void) "" > > +postcopy_request_shared_page(const char *sharer, const char *rb, uin= t64_t rb_offset) "for %s in %s offset 0x%"PRIx64 > > + > > save_xbzrle_page_skipping(void) "" > > save_xbzrle_page_overflow(void) "" > > ram_save_iterate_big_wait(uint64_t milliconds, int iterations) "big = wait: %" PRIu64 " milliseconds, %d iterations" > > -- > > 2.14.3 > > > > >=20 >=20 >=20 > --=20 > Marc-Andr=E9 Lureau -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK