From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: David Gibson <david@gibson.dropbear.id.au>
Cc: aarcange@redhat.com, yamahata@private.email.ne.jp,
quintela@redhat.com, cristian.klein@cs.umu.se,
qemu-devel@nongnu.org, amit.shah@redhat.com,
yanghy@cn.fujitsu.com
Subject: Re: [Qemu-devel] [PATCH v4 44/47] Postcopy; Handle userfault requests
Date: Mon, 5 Jan 2015 17:13:50 +0000 [thread overview]
Message-ID: <20150105171349.GA18097@work-vm> (raw)
In-Reply-To: <20141113032306.GF7291@voom.fritz.box>
* David Gibson (david@gibson.dropbear.id.au) wrote:
> On Fri, Oct 03, 2014 at 06:47:50PM +0100, Dr. David Alan Gilbert (git) wrote:
> > From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> >
> > userfaultfd is a Linux syscall that gives an fd that receives a stream
> > of notifications of accesses to pages marked as MADV_USERFAULT, and
> > allows the program to acknowledge those stalls and tell the accessing
> > thread to carry on.
> >
> > Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
>
> [snip]
> > /*
> > + * Tell the kernel that we've now got some memory it previously asked for.
> > + * Note: We're not allowed to ack a page which wasn't requested.
> > + */
> > +static int ack_userfault(MigrationIncomingState *mis, void *start, size_t len)
> > +{
> > + uint64_t tmp[2];
> > +
> > + /*
> > + * Kernel wants the range that's now safe to access
> > + * Note it always takes 64bit values, even on a 32bit host.
> > + */
> > + tmp[0] = (uint64_t)(uintptr_t)start;
> > + tmp[1] = (uint64_t)(uintptr_t)start + (uint64_t)len;
> > +
> > + if (write(mis->userfault_fd, tmp, 16) != 16) {
> > + int e = errno;
>
> Is an EOF (i.e. write() returns 0) ever possible here? If so errno
> may not have a meaningful value.
I don't think so; I think any !=16 case is an error; however if I understand
correctly the safe thing to do is for me to do an
errno = 0
before the call.
>
> > + if (e == ENOENT) {
> > + /* Kernel said it wasn't waiting - one case where this can
> > + * happen is where two threads triggered the userfault
> > + * and we receive the page and ack it just after we received
> > + * the 2nd request and that ends up deciding it should ack it
> > + * We could optimise it out, but it's rare.
> > + */
> > + /*fprintf(stderr, "ack_userfault: %p/%zx ENOENT\n", start, len); */
> > + return 0;
> > + }
> > + error_report("postcopy_ram: Failed to notify kernel for %p/%zx (%d)",
> > + start, len, e);
> > + return -errno;
Hmm, and made that return -e
> > +/*
> > * Handle faults detected by the USERFAULT markings
> > */
> > static void *postcopy_ram_fault_thread(void *opaque)
> > {
> > MigrationIncomingState *mis = (MigrationIncomingState *)opaque;
> > + void *hostaddr;
> > + int ret;
> > + size_t hostpagesize = getpagesize();
> > + RAMBlock *rb = NULL;
> > + RAMBlock *last_rb = NULL; /* last RAMBlock we sent part of */
> >
> > - fprintf(stderr, "postcopy_ram_fault_thread\n");
> > - /* TODO: In later patch */
> > + DPRINTF("%s", __func__);
> > qemu_sem_post(&mis->fault_thread_sem);
> > - while (1) {
> > - /* TODO: In later patch */
> > - }
> > + while (true) {
> > + PostcopyPMIState old_state, tmp_state;
> > + ram_addr_t rb_offset;
> > + ram_addr_t in_raspace;
> > + unsigned long bitmap_index;
> > + struct pollfd pfd[2];
> > +
> > + /*
> > + * We're mainly waiting for the kernel to give us a faulting HVA,
> > + * however we can be told to quit via userfault_quit_fd which is
> > + * an eventfd
> > + */
> > + pfd[0].fd = mis->userfault_fd;
> > + pfd[0].events = POLLIN;
> > + pfd[0].revents = 0;
> > + pfd[1].fd = mis->userfault_quit_fd;
> > + pfd[1].events = POLLIN; /* Waiting for eventfd to go positive */
> > + pfd[1].revents = 0;
> > +
> > + if (poll(pfd, 2, -1 /* Wait forever */) == -1) {
> > + perror("userfault poll");
> > + break;
> > + }
> >
> > + if (pfd[1].revents) {
> > + DPRINTF("%s got quit event", __func__);
> > + break;
>
> I don't see any cleanup path in the userfault thread. So wouldn't it
> be simpler to just pthread_cancel() it rather than using an extra fd
> for quit notifications.
But it does call functions that take locks (both the pmi and the
return path qemu-file), so I don't feel comfortable just cancelling the
thread.
I guess I could do a pthread_set_cancelstate around the top of the loop
to only allow it to cancel there; is that any better than the fd?
> > @@ -612,11 +814,12 @@ int postcopy_place_page(MigrationIncomingState *mis, void *host, void *from,
> >
> > if (syscall(__NR_remap_anon_pages, host, from, hps, 0) !=
> > getpagesize()) {
> > + int e = errno;
> > perror("remap_anon_pages in postcopy_place_page");
> > fprintf(stderr, "host: %p from: %p pmi=%d\n", host, from,
> > postcopy_pmi_get_state(mis, bitmap_offset));
> >
> > - return -errno;
> > + return -e;
>
> Unrelated change, should probably be folded into the patch which added
> this code.
Thanks, fixed.
Dave
>
> > }
> >
> > tmp_state = postcopy_pmi_get_state(mis, bitmap_offset);
> > @@ -629,7 +832,10 @@ int postcopy_place_page(MigrationIncomingState *mis, void *host, void *from,
> >
> >
> > if (old_state == POSTCOPY_PMI_REQUESTED) {
> > - /* TODO: Notify kernel */
> > + /* Send the kernel the host address that should now be accessible */
> > + DPRINTF("%s: Notifying kernel bitmap_offset=0x%lx host=%p",
> > + __func__, bitmap_offset, host);
> > + return ack_userfault(mis, host, hps);
> > }
> >
> > return 0;
>
> --
> David Gibson | I'll have my music baroque, and my code
> david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
> | _way_ _around_!
> http://www.ozlabs.org/~dgibson
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2015-01-05 17:14 UTC|newest]
Thread overview: 204+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-03 17:47 [Qemu-devel] [PATCH v4 00/47] Postcopy implementation Dr. David Alan Gilbert (git)
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 01/47] QEMUSizedBuffer based QEMUFile Dr. David Alan Gilbert (git)
2014-10-08 2:10 ` zhanghailiang
2014-11-03 0:53 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 02/47] Tests: QEMUSizedBuffer/QEMUBuffer Dr. David Alan Gilbert (git)
2014-11-03 1:02 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 03/47] Start documenting how postcopy works Dr. David Alan Gilbert (git)
2014-11-03 1:31 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 04/47] qemu_ram_foreach_block: pass up error value, and down the ramblock name Dr. David Alan Gilbert (git)
2014-11-03 2:34 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 05/47] improve DPRINTF macros, add to savevm Dr. David Alan Gilbert (git)
2014-11-03 2:35 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 06/47] Add qemu_get_counted_string to read a string prefixed by a count byte Dr. David Alan Gilbert (git)
2014-11-03 2:39 ` David Gibson
2014-11-25 16:13 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 07/47] Create MigrationIncomingState Dr. David Alan Gilbert (git)
2014-11-03 2:45 ` David Gibson
2014-11-04 19:06 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 08/47] socket shutdown Dr. David Alan Gilbert (git)
2014-10-04 18:09 ` Paolo Bonzini
2014-10-07 10:00 ` Dr. David Alan Gilbert
2014-10-07 11:10 ` Paolo Bonzini
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 09/47] Provide runtime Target page information Dr. David Alan Gilbert (git)
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 10/47] Return path: Open a return path on QEMUFile for sockets Dr. David Alan Gilbert (git)
2014-11-03 3:05 ` David Gibson
2014-11-03 19:04 ` Dr. David Alan Gilbert
2014-11-18 4:34 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 11/47] Return path: socket_writev_buffer: Block even on non-blocking fd's Dr. David Alan Gilbert (git)
2014-11-03 3:10 ` David Gibson
2014-11-03 18:59 ` Dr. David Alan Gilbert
2014-11-18 3:54 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 12/47] Handle bi-directional communication for fd migration Dr. David Alan Gilbert (git)
2014-11-03 3:12 ` David Gibson
2014-11-03 13:53 ` Cristian Klein
2014-11-18 3:53 ` David Gibson
2014-11-19 17:27 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 13/47] Migration commands Dr. David Alan Gilbert (git)
2014-11-03 3:14 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 14/47] Return path: Control commands Dr. David Alan Gilbert (git)
2014-10-04 18:08 ` Paolo Bonzini
2014-10-23 16:23 ` Dr. David Alan Gilbert
2014-10-23 20:15 ` Paolo Bonzini
2014-11-03 3:20 ` David Gibson
2014-11-04 18:58 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 15/47] Return path: Send responses from destination to source Dr. David Alan Gilbert (git)
2014-11-03 3:22 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 16/47] Return path: Source handling of return path Dr. David Alan Gilbert (git)
2014-10-04 18:14 ` Paolo Bonzini
2014-10-23 18:00 ` Dr. David Alan Gilbert
2014-10-24 10:04 ` Paolo Bonzini
2014-10-16 8:26 ` zhanghailiang
2014-10-16 8:35 ` Dr. David Alan Gilbert
2014-10-16 9:09 ` zhanghailiang
2014-11-03 3:47 ` David Gibson
2014-11-25 15:44 ` Dr. David Alan Gilbert
2014-11-03 3:46 ` David Gibson
2014-11-03 13:22 ` Dr. David Alan Gilbert
2014-11-18 3:52 ` David Gibson
2014-11-19 17:06 ` Dr. David Alan Gilbert
2014-11-19 21:12 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 17/47] qemu_loadvm errors and debug Dr. David Alan Gilbert (git)
2014-11-03 3:49 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 18/47] ram_debug_dump_bitmap: Dump a migration bitmap as text Dr. David Alan Gilbert (git)
2014-11-03 3:58 ` David Gibson
2014-11-19 17:35 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 19/47] Rework loadvm path for subloops Dr. David Alan Gilbert (git)
2014-10-04 16:46 ` Paolo Bonzini
2014-10-07 8:58 ` Dr. David Alan Gilbert
2014-10-07 10:12 ` Paolo Bonzini
2014-10-07 10:21 ` Dr. David Alan Gilbert
2014-11-03 5:08 ` David Gibson
2014-11-19 17:50 ` Dr. David Alan Gilbert
2014-11-21 6:53 ` David Gibson
2014-12-11 14:47 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 20/47] Add migration-capability boolean for postcopy-ram Dr. David Alan Gilbert (git)
2014-10-06 18:59 ` Eric Blake
2014-10-06 19:07 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 21/47] Add wrappers and handlers for sending/receiving the postcopy-ram migration messages Dr. David Alan Gilbert (git)
2014-11-03 5:51 ` David Gibson
2014-12-17 14:50 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 22/47] QEMU_VM_CMD_PACKAGED: Send a packaged chunk of migration stream Dr. David Alan Gilbert (git)
2014-11-04 1:28 ` David Gibson
2014-11-04 10:19 ` Dr. David Alan Gilbert
2014-11-18 4:36 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 23/47] migrate_init: Call from savevm Dr. David Alan Gilbert (git)
2014-10-08 2:28 ` zhanghailiang
2014-11-04 1:29 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 24/47] Allow savevm handlers to state whether they could go into postcopy Dr. David Alan Gilbert (git)
2014-11-04 1:33 ` David Gibson
2014-11-19 17:53 ` Dr. David Alan Gilbert
2014-11-21 6:58 ` David Gibson
2014-11-25 19:58 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 25/47] postcopy: OS support test Dr. David Alan Gilbert (git)
2014-11-04 1:40 ` David Gibson
2014-11-25 17:34 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 26/47] migrate_start_postcopy: Command to trigger transition to postcopy Dr. David Alan Gilbert (git)
2014-11-04 1:47 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 27/47] MIG_STATE_POSTCOPY_ACTIVE: Add new migration state Dr. David Alan Gilbert (git)
2014-11-04 1:49 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 28/47] qemu_savevm_state_complete: Postcopy changes Dr. David Alan Gilbert (git)
2014-11-04 2:18 ` David Gibson
2014-12-17 16:14 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 29/47] Postcopy page-map-incoming (PMI) structure Dr. David Alan Gilbert (git)
2014-11-04 3:09 ` David Gibson
2014-11-19 18:46 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 30/47] Postcopy: Maintain sentmap and calculate discard Dr. David Alan Gilbert (git)
2014-11-05 6:38 ` David Gibson
2014-12-17 16:48 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 31/47] postcopy: Incoming initialisation Dr. David Alan Gilbert (git)
2014-11-05 6:47 ` David Gibson
2014-12-17 17:21 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 32/47] postcopy: ram_enable_notify to switch on userfault Dr. David Alan Gilbert (git)
2014-10-04 16:42 ` Paolo Bonzini
2014-10-06 19:00 ` Dr. David Alan Gilbert
2014-11-05 6:49 ` David Gibson
2014-11-19 18:59 ` Dr. David Alan Gilbert
2014-11-19 21:17 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 33/47] Postcopy: Postcopy startup in migration thread Dr. David Alan Gilbert (git)
2014-10-04 16:27 ` Paolo Bonzini
2014-11-20 11:45 ` Dr. David Alan Gilbert
2014-11-21 12:01 ` Paolo Bonzini
2014-11-21 12:07 ` Dr. David Alan Gilbert
2014-11-20 17:12 ` Dr. David Alan Gilbert
2014-11-20 17:19 ` Paolo Bonzini
2014-11-24 18:26 ` Dr. David Alan Gilbert
2014-11-10 6:05 ` David Gibson
2015-01-05 16:06 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 34/47] Postcopy: Create a fault handler thread before marking the ram as userfault Dr. David Alan Gilbert (git)
2014-11-10 6:10 ` David Gibson
2014-11-19 18:56 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 35/47] Page request: Add MIG_RPCOMM_REQPAGES reverse command Dr. David Alan Gilbert (git)
2014-11-10 6:19 ` David Gibson
2014-11-19 20:01 ` Dr. David Alan Gilbert
2014-11-19 21:48 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 36/47] Page request: Process incoming page request Dr. David Alan Gilbert (git)
2014-10-08 2:31 ` zhanghailiang
2014-10-08 7:49 ` Dr. David Alan Gilbert
2014-10-08 8:07 ` Paolo Bonzini
2014-10-08 8:10 ` zhanghailiang
2014-10-08 8:18 ` Dr. David Alan Gilbert
2014-11-10 6:31 ` David Gibson
2014-11-17 19:07 ` Dr. David Alan Gilbert
2014-11-18 4:38 ` David Gibson
2014-11-19 19:37 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 37/47] Page request: Consume pages off the post-copy queue Dr. David Alan Gilbert (git)
2014-10-04 18:04 ` Paolo Bonzini
2014-10-07 11:35 ` Dr. David Alan Gilbert
2014-11-11 1:13 ` David Gibson
2015-01-14 20:13 ` Dr. David Alan Gilbert
2015-01-27 4:38 ` David Gibson
2015-01-27 9:40 ` Dr. David Alan Gilbert
2015-01-28 5:33 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 38/47] Add assertion to check migration_dirty_pages Dr. David Alan Gilbert (git)
2014-10-04 18:32 ` Paolo Bonzini
2014-10-06 18:51 ` Dr. David Alan Gilbert
2014-10-06 20:30 ` Paolo Bonzini
2014-11-11 1:14 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 39/47] postcopy_ram.c: place_page and helpers Dr. David Alan Gilbert (git)
2014-11-11 1:39 ` David Gibson
2015-01-15 18:14 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 40/47] Postcopy: Use helpers to map pages during migration Dr. David Alan Gilbert (git)
2014-11-13 2:53 ` David Gibson
2014-11-25 18:14 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 41/47] qemu_ram_block_from_host Dr. David Alan Gilbert (git)
2014-11-13 2:59 ` David Gibson
2014-11-25 18:55 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 42/47] Don't sync dirty bitmaps in postcopy Dr. David Alan Gilbert (git)
2014-11-13 3:01 ` David Gibson
2014-11-25 16:25 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 43/47] Host page!=target page: Cleanup bitmaps Dr. David Alan Gilbert (git)
2014-11-13 3:10 ` David Gibson
2014-12-17 18:21 ` Dr. David Alan Gilbert
2015-01-27 4:50 ` David Gibson
2015-01-27 10:04 ` Dr. David Alan Gilbert
2015-01-28 5:36 ` David Gibson
2015-01-27 10:20 ` Peter Maydell
2015-01-27 11:50 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 44/47] Postcopy; Handle userfault requests Dr. David Alan Gilbert (git)
2014-11-13 3:23 ` David Gibson
2015-01-05 17:13 ` Dr. David Alan Gilbert [this message]
2015-01-27 4:33 ` David Gibson
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 45/47] Start up a postcopy/listener thread ready for incoming page data Dr. David Alan Gilbert (git)
2014-11-13 3:29 ` David Gibson
2014-11-19 19:40 ` Dr. David Alan Gilbert
2014-11-21 8:36 ` David Gibson
2014-11-21 10:17 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 46/47] postcopy: Wire up loadvm_postcopy_ram_handle_{run, end} commands Dr. David Alan Gilbert (git)
2014-10-04 17:51 ` Paolo Bonzini
2014-10-23 12:18 ` Dr. David Alan Gilbert
2014-10-03 17:47 ` [Qemu-devel] [PATCH v4 47/47] End of migration for postcopy Dr. David Alan Gilbert (git)
2014-10-04 17:49 ` Paolo Bonzini
2014-10-23 14:24 ` Dr. David Alan Gilbert
2014-10-04 18:31 ` Paolo Bonzini
2014-10-07 10:29 ` Dr. David Alan Gilbert
2014-10-07 11:12 ` Paolo Bonzini
2014-10-03 19:21 ` [Qemu-devel] [PATCH v4 00/47] Postcopy implementation Dr. David Alan Gilbert
2014-10-07 2:27 ` Cristian Klein
2014-10-07 8:12 ` Dr. David Alan Gilbert
2014-10-08 8:36 ` Cristian Klein
2014-11-21 3:48 ` zhanghailiang
2014-11-21 10:14 ` Dr. David Alan Gilbert
2014-11-24 8:10 ` zhanghailiang
2014-11-21 18:56 ` Andrea Arcangeli
2014-11-24 8:25 ` zhanghailiang
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=20150105171349.GA18097@work-vm \
--to=dgilbert@redhat.com \
--cc=aarcange@redhat.com \
--cc=amit.shah@redhat.com \
--cc=cristian.klein@cs.umu.se \
--cc=david@gibson.dropbear.id.au \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=yamahata@private.email.ne.jp \
--cc=yanghy@cn.fujitsu.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).