hyperv/storvsc: Fix SRB length setting.
This fixes disk discovery issue on WS2008R2 Hyper-V, which plagued
us since 10.2-release.
Reported by: many
Sponsored by: Microsoft
hyperv/storvsc: Deliver CAM_SEL_TIMEOUT upon SRB status error.
SRB status is set to 0x20 by the hypervisor, if the specified LUN is
unaccessible, and even worse the INQUIRY response will not be set by
the hypervisor at all under this situation. Additionally, SRB status
is 0x20 too, for TUR on an unaccessible LUN.
Deliver CAM_SEL_TIMEOUT to CAM upon SRB status errors as suggested by
Scott Long, other values seems improper.
This commit fixes the Hyper-V disk hotplug support.
Submitted by: Hongjiang Zhang <honzhan microsoft com>
Sponsored by: Microsoft
Differential Revision: https://reviews.freebsd.org/D7521
hyperv/storvsc: Claim SPC-3 conformance, thus enable UNMAP support
The Hyper-V on pre-win10 systems will only report SPC-2 conformance,
but it actually conforms to SPC-3. The INQUIRY response is adjusted
to propagate the SPC-3 version information to CAM.
Submitted by: Hongjiang Zhang <honzhan microsoft com>
Sponsored by: Microsoft
Differential Revision: https://reviews.freebsd.org/D7405
302541
hyperv/stor: Fix the INQUIRY checks
Don't check the area that the host has not filled.
PR: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=209443
PR: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=210425
Submitted by: Hongjiang Zhang <honzhan microsoft com>
Reviewed by: sephe, Dexuan Cui <decui microsoft com>
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6955
302605
hyperv/stor: Save the response status and xfer length properly.
The current command response handling discards status and xfer
length unconditionally, so that all of the commands would be
considered successful, even if errors happened. When errors
really happens, this causes all kinds of wiredness, since the
buffer will not be filled on the host side and sense data will
be ignored.
Most of the time, errors do not happen, however, error does
happen for the request sent immediately after the disk resizing.
Discarding the SCSI status (SCSI_STATUS_CHECK_COND) and sense
data (capacity changes) prevents the disk resizing from working
properly.
This commit saves the response status and xfer length properly
for later use.
Submitted by: Dexuan Cui <decui microsoft com>
Noticed by: sephe
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D7181
300987
hyperv/et: Fix STIMER0 operations.
- Make sure that STIMER0 is disabled before writting to it, since
writing to an enabled STIMER will result in undefined behaviour.
- It is unnecessary to reconfigure STIMER0 upon each et_start().
- Make sure that MSR_HV_REF_TIME_COUNT will not return 0, since
writing 0 to STIMER_COUNT will disable the target STIMER.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6573
300988
hyperv/vmbus: Move SINT settings to vmbus_var.h
While I'm here remove the event timer's dependency on hv_vmbus_priv.h
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6574
300989
hyperv/et: Make sure only one event timer will be registered
This nullifies the need to use softc.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6591
300992
hyperv: Move timer frequency definition to common place.
And cleanup event timer period settings.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6597
300993
hyperv/et: Device renaming; consistent w/ other Hyper-V utils
While I'm here, prefix function names w/ vmbus, since unlike Hyper-V
timecounter, Hyper-V event timer will not work w/o vmbus.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6598
300994
hyperv/et: Allow Hyper-V event timer be disabled
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6599
301009
hyperv/vmbus: Process event timer before checking events
And update comment.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6600
hyperv/vmbus: Allocate/setup IDT vector after all ISR resources are ready
And release IDT vector before releasing ISR resources on interrupt
teardown path. We still have some work to do on the interrupt tearing
down path.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6519
300576
hyperv/vmbus: Free message taskqueue during interrupt teardown
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6506
300644
hyperv/vmbus: Check hyperv_dmamem_alloc return value
Though it is highly unlikely this function would fail w/ BUS_DMA_WAITOK,
we had better to check its return value; better safe then sorry here.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6518
hyperv/vmbus: Factor out functions for vmbus interrupt set/teardown
This paves way for further cleanup and fix.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6505
hyperv/vmbus: Git rid of sc version of pcpu data extraction macro
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6503
hyperv/vmbus: Use busdma(9) for messages and event flags
And
- Move message and event flags to vmbus_softc per-cpu data.
- Get rid of hv_setup_arg, which serves no purpose now.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6502
hyperv/vmbus: Move IDT vector to vmbus_softc
Prepare to get rid of the hv_setup_arg.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6449
300480
hyperv: Move Hypercall setup to an early place.
It does not belong to the vmbus.
While I'm here rework the Hypercall setup, e.g. use busdma(9)
and avoid bit fields.
Discussed with: Jun Su <junsu microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6445
300481
hyperv/vmbus: Declare Synic message and event w/ proper types
Avoid ugly casts.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6446
300486
hyperv/vmbus: Get rid of vmbus_devp
While I'm here, nuke useless print in vmbus_attach().
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6447
300122
hyperv: Set vm_guest to VM_GUEST_VM, if hypervisor is not Hyper-V
Reviewed by: kib
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6412
300123
hyperv/vmbus: Staticize vmbus_devclass
Reviewed by: Jun Su <junsu microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6414
300124
hyperv/vmbus: Reindent and cleanup devmethods.
While I'm here, use DEVMETHOD_END.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6415
300126
hyperv/vmbus: Fix SYSINIT function prototype and usage.
Reviewed by: Jun Su <junsu microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6416
300127
hyperv/vmbus: Minor function definition style fixup
Reviewed by: Jun Su <junsu microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6417
300129
hyperv/vmbus: Use consistent device description as other devices
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6418
300455
hyperv: Move guest id setup to early place
And
- Rework the guest id composition.
- Nuke useless saved guest_id.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6430
299927
hyperv/vmbus: Use atomic_testandclear
Prepare to use unsigned long for event channel bit array.
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6382
300101
hyperv/vmbus: Use unsigned long for event bits.
And move base channel id calculation out of inner loop. This prepares
for more event processing optimization.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6384
300102
hyperv/vmbus: Reduce the # of event loops by recording event flag count
Use vmbus softc to save vmbus per-cpu data. More stuffs will be moved
into vmbus softc.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6403
300105
hyperv/vmbus: Use atomic swap and flsl to process event flags
Greatly reduce the locked instructions and reduce number of inner loops.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6404
300107
hyperv/vmbus: Avoid two unnecessary protocol checks on isr handling path
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6405
299889
hyperv/vmbus: Simplify event processing
While I'm here, remove useless comment and unnecessary return.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6332
299890
hyperv/vmbus: Simplify event processing
For channel0, it will never be processed on event handling path,
so there is no need to install it. After skipping in the channel0
installation, we could discard the channel0 check on event
handling hot code path.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6333
299892
hyperv/vmbus: Fix event processing loop indentation.
No functional changes.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6334
298618
hyperv/stor: Set description properly in probe devmethod
MFC after: 1 week
Sponsored by: Microsoft OSTC
298620
hyperv/hn: Change description to "Hyper-V Network Interface"
This is consistent w/ other Hyper-V devices.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298688
hyperv/hn: Restart sending earlier once we gathered some free TX descs
This greatly reduces the oqdrops under heavy workload.
For TCP send/recv test (10K concurrent connections):
oqdrops is reduced by 17% on sending side, and 57% on receiving side.
For nginx-1.8/wrk-4 1KB object test (10K concurrent connections,
4 requests/connection):
oqdrops is reduced by 44% on nginx side, and 10% on wrk side.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298689
hyperv/hn: Remove unapplied comment.
Chimney sending buffers are shared across channels.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298569
hyperv/channel: Remove the unnecessary 'new' flag
MFC after: 1 week
Sponsored by: Microsoft OSTC
298570
hyperv/channel: Minor style changes; no functional changes.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298571
hyperv/channel: Remove unapplied comment
MFC after: 1 week
Sponsored by: Microsoft OSTC
298572
hyperv/channel: Log a warning about duplicated primary channel offer
MFC after: 1 week
Sponsored by: Microsoft OSTC
298574
hyperv/channel: Add functions to synchronize sub-channel offers
MFC after: 1 week
Sponsored by: Microsoft OSTC
298575
hyperv/hn: Synchronize sub-channel offers
MFC after: 1 week
Sponsored by: Microsoft OSTC
298613
hyperv/stor: Remove the useless hs_open_multi_channel
This fixes the sub-channel offer race after Hyper-V device probe/attach
is moved to vmbus SYSINIT/attach.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298614
hyperv/stor: Synchronize sub-channel offers
MFC after: 1 week
Sponsored by: Microsoft OSTC
298615
hyperv/hn: Avoid sub-channel creation callback.
Since the sub-channel offers are synchronized, we can do our own
channel setup without using the sub-channel creation callback.
This paves the way to whack the sub-channel creation callback.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298616
hyperv/stor: Avoid sub-channel creation callback.
Since the sub-channel offers are synchronized, we can do our own
channel setup without using the sub-channel creation callback.
This paves the way to whack the sub-channel creation callback.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298617
hyperv/channel: Git rid of the sub-channel creation callback
It is no longer used.
MFC after: 1 week
Sponsored by: Microsoft OSTC
298449
hyperv/et: Make Hyper-V event timer a device.
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: sephe, Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5957
298568
hyperv/et: Strip extra white space in function name
Reported by: Sascha Wildner <saw online de>
MFC after: 1 week
Sponsored by: Microsoft OSTC
hyperv: Update copyright to 2016 for the files Microsoft changed in 2016
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D6039
hyperv/vmbus: Make device probe/attach synchronous w/ vmbus attach/SYSINIT
Discussed with: Jun Su <junsu microsoft com>, Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
298041
hyperv: No need to zero out softc
MFC after: 1 week
Sponsored by: Microsoft OSTC
298259
hyperv: Remove two assign-only local variables
Submitted by: Jun Su <junsu microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
297931
Expose doreti as a global symbol on amd64 and i386.
doreti provides the common code path for returning from interrupt
andlers on x86. Exposing doreti as a global symbol allows kernel
modules to include low-level interrupt handlers instead of requiring
all low-level handlers to be statically compiled into the kernel.
Submitted by: Howard Su <howard0su@gmail.com>
Reviewed by: kib
298022
hyperv: Deprecate HYPERV option by moving Hyper-V IDT vector into vmbus
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: jhb, kib, sephe
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5910
298039
hyperv/vmbus: Put multi-channel offer logging under bootverbose
Suggested by: Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
298042
hyperv/hn: Hide ring to channel linkage message under bootverbose
Suggested by: Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
297809
hyperv/hn: Cap default # of rings to 8.
8 gives the best performance in both Azure and local Hyper-V on both
10Ge and 40Ge. More rings are still allowed by manual configuration.
Reviewed by: Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5879
297810
hyperv/hn: By default enable multiple TX/RX rings, aka vRSS.
Reviewed by: Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5880
297811
hyperv/hn: Remove unnecessary NULL checks
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5905
297636
hyperv: Typo in r297634
Noticed by: hiren
MFC after: 1 week
Sponsored by: Microsoft OSTC
297800
hyperv/vmbus: Nuke unused function
MFC after: 1 week
Sponsored by: Microsoft OSTC
297801
hyperv/vmbus: Get rid of max_leaf detection; this is actually not used.
It will be replaced by a new one.
MFC after: 1 week
Sponsored by: Microsoft OSTC
297805
hyperv: Break long line
MFC after: 1 week
Sponsored by: Microsoft OSTC
297807
hyperv: Print more features
And add comment about the MSR features.
MFC after: 1 week
Sponsored by: Microsoft OSTC
297808
hyperv: Define macro for Hyper-V interface
Suggested by: rpokala
MFC after: 1 week
Sponsored by: Microsoft OSTC
297180
hyperv/hn: Reduce TCP segment aggregation limit for multiple RX rings
This mainly used to improve ACK timeliness when multiple RX rings
are enabled.
This value gives the best performance in both Azure and Hyper-V
environment, w/ both 10Ge and 40Ge using non-{INVARIANTS,WITNESS}
kernel.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5691
297181
hyperv/hn: Factor out hn_set_lro_lenlim()
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5692
297182
hyperv/hn: When short of mbufs on the RX path, don't spam the console.
Instead, increase the IQDROPS counter.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5693
297142
hyperv: Factor out snprinf_hv_guid()
Submitted by: Ju Sun <junsu microsoft com>
Reviewed by: Dexuan Cui <decui microsoft com>, sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5651
297143
hyperv/vmbus: Implement bus_child_pnpinfo_str method
Submitted by: Jun Su <junsu microsoft com>
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5669
297176
hyperv/evttimer: Use an independent message slot so that it can work
Using the same message slot as the other types of the messages has
the side effect that the event timer message could be deferred to
the swi threads to run (lacking of trapframe and the original code
didn't even handle that, so the event timer was actually broken).
As of this commit we use an independent message slot for event timer,
so that we could handle all of event timer messages in the interrupt
handler directly. Note, the message slot for event timer is still
bind to the same interrupt vector as the other types of messages.
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: sephe
Discussed with: Jun Su <junsu microsoft com>, Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5696
297177
hyperv/vmbus: Use taskqueue_fast for non-performance critical messages
This gets rid of the per-cpu SWIs.
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: Dexuan Cui <decui microsoft com>, sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5215
297178
hyperv/vmbus: Remove NULL check for taskqueue_create_fast(M_WAITOK)
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: Dexuan Cui <decui microsoft com>, sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5215
297221
hyperv/vmbus: Create per-cpu fast taskqueue for msg handling
Using one taskqueue does not work, since the EOM MSR must be written
on the msg's owner CPU.
Noticed by: Jun Su <junsu microsoft com>
Discussed with: Jun Su <junsu microsoft com>, Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
296379
hyperv/hn: Add multiple channel support, a.k.a. vRSS
Each channel contains one RX ring and one TX ring. And we
try to distribute the channels to different evenly.
Note: Currently we don't have enough information to extract
the RSS type and RSS hash value from the received packets.
This greatly improves the TX/RX performance for 8 virtual CPU
Hyper-V over 10Ge: it can max out 10Ge for TCP when multiple
RX/TX rings are enabled.
This almost doubles the TX/RX performance for locally connected
Hyper-Vs: was 6Gbps w/ 128 TCP streams, now 11Gbps w/ multiple
RX/TX rings enabled.
It is not enabled by default; it will be switched on after more
tests.
Collaborated with: Hongjiang Zhang <honzhan microsoft com>
MFC after: 2 week
Sponsored by: Microsoft OSTC
296380
hyperv/hn: Pass channel to send done callbacks.
Mainly to strigent the data packet send done check.
MFC after: 2 weeks
Sponsored by: Microsoft OSTC
296381
hyperv/hn: Add per-TX ring stats for # of transmitted packets
MFC after: 2 weeks
Sponsored by: Microsoft OSTC
296593
hyperv/hn: Move if_initname to an earlier place
So that functions shared w/ attach path could use if_printf().
While I'm here, remove unnecessary if_dunit and if_dname assignment.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5576
296594
hyperv/hn: Factor out hn_channel_attach
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5577
296595
hyperv/hn: Make the # of TX rings configurable.
Rename the tunables to avoid confusion.
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5578
297802
hyperv: Identify Hyper-V features and recommends properly
Features bits will be used to detect devices, e.g. timers, which
do not have corresponding event channels.
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: sephe, Dexuan Cui <decui microsoft com>
Rearranged by: sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
279803(297481)
hyperv: Register Hyper-V timer early enough for TSC freq calibration
The i8254 simulation in Hyper-V is kinda broken and is not available
in Generation 2 Hyper-V VMs, so Hyper-V timer must be registered early
enough so that it can be used to do the TSC freq calibration.
This fixes the notorious warning like this:
calcru: runtime went backwards from 50 usec to 25 usec for pid 0 (kernel)
Submitted by: Dexuan Cui <decui microsoft com>
Reviewed by: kib, sephe
Tested by: kib, sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5778
=================
hyperv: Resurrect r297481
This time we make sure that the TIME_REF_COUNT MSR exists.
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: sephe, Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
297804
hyperv: Declare hyperv_{features,recommends} properly
MFC after: 1 week
Sponsored by: Microsoft OSTC
hyperv/vmbus: Use default mtx for channel message queue
First of all sema_post() can't be called w/ spinlock, and the channel
message queue processing is not on hot code path, i.e. spinlock is not
necessary.
Submitted by: Jun Su <junsu microsoft com>
Reviewed by: sephe, Dexuan Cui <decui microsoft com>
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5812
hyperv/vmbus: use a better retry method in hv_vmbus_post_message()
Most often, hv_vmbus_post_message() doesn't fail. However, it fails
intermittently when GPADLs of large shared memory is to be established
with the host, e.g. on the hn(4) attach path: a GPADL of 15MB sendbuf
is created, for which lots of messages will be flooded to the host.
The host side tries to throttle the message rate by returning
HV_STATUS_INSUFFICIENT_BUFFERS.
Before this commit, we do several retries for failed messages, but the
delay between each retry is pretty/too low, which will cause sporadic
message posting failure. We now use large delay (>=1ms) between each
retry to fix the message posting failure.
Submitted by: Dexuan Cui <decui microsoft com>
Reviewed by: sephe
MFC after: 1 week
Sponsored by: Microsoft OSTC
Differential Revision: https://reviews.freebsd.org/D5715
