* Alexey Perevalov (a.pereva...@samsung.com) wrote:
> Initially int64_t was used, but on PowerPC architecture,
> clang doesn't have atomic_*_8 function, so it produces
> link time error.
>
> QEMU is working with time as with 64bit value, but by
> fact 32 bit is enough with CLOCK_REALTIME. In this case
> blocktime will keep only 1200 hours time interval.
>
> Signed-off-by: Alexey Perevalov <a.pereva...@samsung.com>
> ---
> hmp.c | 4 ++--
> migration/postcopy-ram.c | 47 ++++++++++++++++++++++++++++++-----------------
> migration/trace-events | 4 ++--
> qapi/migration.json | 4 ++--
> 4 files changed, 36 insertions(+), 23 deletions(-)
>
> diff --git a/hmp.c b/hmp.c
> index c6bab53..3c376b3 100644
> --- a/hmp.c
> +++ b/hmp.c
> @@ -265,7 +265,7 @@ void hmp_info_migrate(Monitor *mon, const QDict *qdict)
> }
>
> if (info->has_postcopy_blocktime) {
> - monitor_printf(mon, "postcopy blocktime: %" PRId64 "\n",
> + monitor_printf(mon, "postcopy blocktime: %u\n",
> info->postcopy_blocktime);
> }
>
> @@ -273,7 +273,7 @@ void hmp_info_migrate(Monitor *mon, const QDict *qdict)
> Visitor *v;
> char *str;
> v = string_output_visitor_new(false, &str);
> - visit_type_int64List(v, NULL, &info->postcopy_vcpu_blocktime, NULL);
> + visit_type_uint32List(v, NULL, &info->postcopy_vcpu_blocktime, NULL);
> visit_complete(v, &str);
> monitor_printf(mon, "postcopy vcpu blocktime: %s\n", str);
> g_free(str);
> diff --git a/migration/postcopy-ram.c b/migration/postcopy-ram.c
> index 7814da5..ce91de8 100644
> --- a/migration/postcopy-ram.c
> +++ b/migration/postcopy-ram.c
> @@ -63,14 +63,14 @@ struct PostcopyDiscardState {
>
> typedef struct PostcopyBlocktimeContext {
> /* time when page fault initiated per vCPU */
> - int64_t *page_fault_vcpu_time;
> + uint32_t *page_fault_vcpu_time;
> /* page address per vCPU */
> uintptr_t *vcpu_addr;
> - int64_t total_blocktime;
> + uint32_t total_blocktime;
> /* blocktime per vCPU */
> - int64_t *vcpu_blocktime;
> + uint32_t *vcpu_blocktime;
> /* point in time when last page fault was initiated */
> - int64_t last_begin;
> + uint32_t last_begin;
> /* number of vCPU are suspended */
> int smp_cpus_down;
>
> @@ -99,22 +99,22 @@ static void migration_exit_cb(Notifier *n, void *data)
> static struct PostcopyBlocktimeContext *blocktime_context_new(void)
> {
> PostcopyBlocktimeContext *ctx = g_new0(PostcopyBlocktimeContext, 1);
> - ctx->page_fault_vcpu_time = g_new0(int64_t, smp_cpus);
> + ctx->page_fault_vcpu_time = g_new0(uint32_t, smp_cpus);
> ctx->vcpu_addr = g_new0(uintptr_t, smp_cpus);
> - ctx->vcpu_blocktime = g_new0(int64_t, smp_cpus);
> + ctx->vcpu_blocktime = g_new0(uint32_t, smp_cpus);
>
> ctx->exit_notifier.notify = migration_exit_cb;
> qemu_add_exit_notifier(&ctx->exit_notifier);
> return ctx;
> }
>
> -static int64List *get_vcpu_blocktime_list(PostcopyBlocktimeContext *ctx)
> +static uint32List *get_vcpu_blocktime_list(PostcopyBlocktimeContext *ctx)
> {
> - int64List *list = NULL, *entry = NULL;
> + uint32List *list = NULL, *entry = NULL;
> int i;
>
> for (i = smp_cpus - 1; i >= 0; i--) {
> - entry = g_new0(int64List, 1);
> + entry = g_new0(uint32List, 1);
> entry->value = ctx->vcpu_blocktime[i];
> entry->next = list;
> list = entry;
> @@ -145,7 +145,7 @@ void
> fill_destination_postcopy_migration_info(MigrationInfo *info)
> info->postcopy_vcpu_blocktime = get_vcpu_blocktime_list(bc);
> }
>
> -static uint64_t get_postcopy_total_blocktime(void)
> +static uint32_t get_postcopy_total_blocktime(void)
> {
> MigrationIncomingState *mis = migration_incoming_get_current();
> PostcopyBlocktimeContext *bc = mis->blocktime_ctx;
> @@ -619,6 +619,16 @@ static int get_mem_fault_cpu_index(uint32_t pid)
> return -1;
> }
>
> +static uint32_t get_least_significant_part(int64_t value)
> +{
> + unsigned char *t = (unsigned char *)&value;
> +#if defined(HOST_WORDS_BIGENDIAN)
> + return t[4] << 24 | t[5] << 16 | t[6] << 8 | t[7] << 0;
> +#else
> + return t[0] << 0 | t[1] << 8 | t[2] << 16 | t[3] << 24;
> +#endif /* HOST_WORDS_BIGENDIAN */
> +}
This doesn't feel right.
Firstly, we're doing a check for the magic value of read_vcpu_time==0 in
mark_postcopy_blocktime_end - so we have to be careful not to hit it.
Just masking the bottom 32bits of time means we've got a (rare) chance
of hitting that; but we've got a much less rare change of hitting
the case where one of the measurements happens after the roll-over
of the bottom 32bits.
If you stored a time at the start of the postcopy and just
subtracted that from 'now' you're probably OK though.
Anding a 64bit value with 0xffffffff and casting is probably much
easier than shifting etc.
Dave
> /*
> * This function is being called when pagefault occurs. It
> * tracks down vCPU blocking time.
> @@ -634,6 +644,7 @@ static void mark_postcopy_blocktime_begin(uintptr_t addr,
> uint32_t ptid,
> MigrationIncomingState *mis = migration_incoming_get_current();
> PostcopyBlocktimeContext *dc = mis->blocktime_ctx;
> int64_t now_ms;
> + uint32_t least_now;
>
> if (!dc || ptid == 0) {
> return;
> @@ -644,13 +655,14 @@ static void mark_postcopy_blocktime_begin(uintptr_t
> addr, uint32_t ptid,
> }
>
> now_ms = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
> + least_now = get_least_significant_part(now_ms);
> if (dc->vcpu_addr[cpu] == 0) {
> atomic_inc(&dc->smp_cpus_down);
> }
>
> - atomic_xchg__nocheck(&dc->last_begin, now_ms);
> - atomic_xchg__nocheck(&dc->page_fault_vcpu_time[cpu], now_ms);
> - atomic_xchg__nocheck(&dc->vcpu_addr[cpu], addr);
> + atomic_xchg(&dc->last_begin, least_now);
> + atomic_xchg(&dc->page_fault_vcpu_time[cpu], least_now);
> + atomic_xchg(&dc->vcpu_addr[cpu], addr);
>
> /* check it here, not at the begining of the function,
> * due to, check could accur early than bitmap_set in
> @@ -699,20 +711,21 @@ static void mark_postcopy_blocktime_end(uintptr_t addr)
> int i, affected_cpu = 0;
> int64_t now_ms;
> bool vcpu_total_blocktime = false;
> - int64_t read_vcpu_time;
> + uint32_t read_vcpu_time, least_now;
>
> if (!dc) {
> return;
> }
>
> now_ms = qemu_clock_get_ms(QEMU_CLOCK_REALTIME);
> + least_now = get_least_significant_part(now_ms);
>
> /* lookup cpu, to clear it,
> * that algorithm looks straighforward, but it's not
> * optimal, more optimal algorithm is keeping tree or hash
> * where key is address value is a list of */
> for (i = 0; i < smp_cpus; i++) {
> - uint64_t vcpu_blocktime = 0;
> + uint32_t vcpu_blocktime = 0;
>
> read_vcpu_time = atomic_fetch_add(&dc->page_fault_vcpu_time[i], 0);
> if (atomic_fetch_add(&dc->vcpu_addr[i], 0) != addr ||
> @@ -720,7 +733,7 @@ static void mark_postcopy_blocktime_end(uintptr_t addr)
> continue;
> }
> atomic_xchg__nocheck(&dc->vcpu_addr[i], 0);
> - vcpu_blocktime = now_ms - read_vcpu_time;
> + vcpu_blocktime = least_now - read_vcpu_time;
> affected_cpu += 1;
> /* we need to know is that mark_postcopy_end was due to
> * faulted page, another possible case it's prefetched
> @@ -735,7 +748,7 @@ static void mark_postcopy_blocktime_end(uintptr_t addr)
>
> atomic_sub(&dc->smp_cpus_down, affected_cpu);
> if (vcpu_total_blocktime) {
> - dc->total_blocktime += now_ms - atomic_fetch_add(&dc->last_begin, 0);
> + dc->total_blocktime += least_now - atomic_fetch_add(&dc->last_begin,
> 0);
> }
> trace_mark_postcopy_blocktime_end(addr, dc, dc->total_blocktime,
> affected_cpu);
> diff --git a/migration/trace-events b/migration/trace-events
> index 141e773..0defbc3 100644
> --- a/migration/trace-events
> +++ b/migration/trace-events
> @@ -115,8 +115,8 @@ process_incoming_migration_co_end(int ret, int ps)
> "ret=%d postcopy-state=%d"
> process_incoming_migration_co_postcopy_end_main(void) ""
> migration_set_incoming_channel(void *ioc, const char *ioctype) "ioc=%p
> ioctype=%s"
> migration_set_outgoing_channel(void *ioc, const char *ioctype, const char
> *hostname) "ioc=%p ioctype=%s hostname=%s"
> -mark_postcopy_blocktime_begin(uint64_t addr, void *dd, int64_t time, int
> cpu, int received) "addr: 0x%" PRIx64 ", dd: %p, time: %" PRId64 ", cpu: %d,
> already_received: %d"
> -mark_postcopy_blocktime_end(uint64_t addr, void *dd, int64_t time, int
> affected_cpu) "addr: 0x%" PRIx64 ", dd: %p, time: %" PRId64 ", affected_cpu:
> %d"
> +mark_postcopy_blocktime_begin(uint64_t addr, void *dd, uint32_t time, int
> cpu, int received) "addr: 0x%" PRIx64 ", dd: %p, time: %u, cpu: %d,
> already_received: %d"
> +mark_postcopy_blocktime_end(uint64_t addr, void *dd, uint32_t time, int
> affected_cpu) "addr: 0x%" PRIx64 ", dd: %p, time: %u, affected_cpu: %d"
>
> # migration/rdma.c
> qemu_rdma_accept_incoming_migration(void) ""
> diff --git a/qapi/migration.json b/qapi/migration.json
> index 70e7b67..ee55d7c 100644
> --- a/qapi/migration.json
> +++ b/qapi/migration.json
> @@ -175,8 +175,8 @@
> '*setup-time': 'int',
> '*cpu-throttle-percentage': 'int',
> '*error-desc': 'str',
> - '*postcopy-blocktime' : 'int64',
> - '*postcopy-vcpu-blocktime': ['int64']} }
> + '*postcopy-blocktime' : 'uint32',
> + '*postcopy-vcpu-blocktime': ['uint32']} }
>
> ##
> # @query-migrate:
> --
> 2.7.4
>
--
Dr. David Alan Gilbert / dgilb...@redhat.com / Manchester, UK
