Allocate ring buffer memory from the NUMA node assigned to the channel.
Since this is a performance and not a correctness issue, if the node specific
allocation were to fail, fall back and allocate without specifying the node.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Channels/sub-channels can be affinitized to VCPUs in the guest. Implement
this affinity in a way that is NUMA aware. The current protocol distributed
the primary channels uniformly across all available CPUs. The new protocol
is NUMA aware: primary channels are distributed across the available NUMA
nodes while the sub-channels within a primary channel are distributed amongst
CPUs within the NUMA node assigned to the primary channel.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Map target_cpu to target_vcpu using the mapping table.
We should use the mapping table to transform guest CPU ID to VP Index
as is done for the non-performance critical channels.
While the value CPU 0 is special and will
map to VP index 0, it is good to be consistent.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Memory notifiers are being executed in a sequential order and when one of
them fails returning something different from NOTIFY_OK the remainder of
the notification chain is not being executed. When a memory block is being
onlined in online_pages() we do memory_notify(MEM_GOING_ONLINE, ) and if
one of the notifiers in the chain fails we end up doing
memory_notify(MEM_CANCEL_ONLINE, ) so it is possible for a notifier to see
MEM_CANCEL_ONLINE without seeing the corresponding MEM_GOING_ONLINE event.
E.g. when CONFIG_KASAN is enabled the kasan_mem_notifier() is being used
to prevent memory hotplug, it returns NOTIFY_BAD for all MEM_GOING_ONLINE
events. As kasan_mem_notifier() comes before the hv_memory_notifier() in
the notification chain we don't see the MEM_GOING_ONLINE event and we do
not take the ha_region_mutex. We, however, see the MEM_CANCEL_ONLINE event
and unconditionally try to release the lock, the following is observed:
[ 110.850927] =====================================
[ 110.850927] [ BUG: bad unlock balance detected! ]
[ 110.850927] 4.1.0-rc3_bugxxxxxxx_test_xxxx #595 Not tainted
[ 110.850927] -------------------------------------
[ 110.850927] systemd-udevd/920 is trying to release lock
(&dm_device.ha_region_mutex) at:
[ 110.850927] [<ffffffff81acda0e>] mutex_unlock+0xe/0x10
[ 110.850927] but there are no more locks to release!
At the same time we can have the ha_region_mutex taken when we get the
MEM_CANCEL_ONLINE event in case one of the memory notifiers after the
hv_memory_notifier() in the notification chain failed so we need to add
the mutex_is_locked() check. In case of MEM_ONLINE we are always supposed
to have the mutex locked.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Add support for Windows 10.
Signed-off-by: Keith Mange <keith.mange@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Primary channels are distributed evenly across all vcpus we have. When the host
asks us to create subchannels it usually makes us num_cpus-1 offers and we are
supposed to distribute the work evenly among the channel itself and all its
subchannels. Make sure they are all assigned to different vcpus.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We need to call init_vp_index() after we added the channel to the appropriate
list (global or subchannel) to be able to use this information when assigning
the channel to the particular vcpu. To do so we need to move a couple of
functions around. The only real change is the init_vp_index() call. This is a
small refactoring without a functional change.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It is unlikely that that host will ask us to close only one subchannel for a
device but let's be consistent. Do both num_sc++ and num_sc-- with
channel->lock to be on the safe side.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case there was an error reported in the response to the CHANNELMSG_OPENCHANNEL
call we need to do the cleanup as a vmbus_open() user won't be doing it after
receiving an error. The cleanup should be done on all failure paths. We also need
to avoid returning open_info->response.open_result.status as the return value as
all other errors we return from vmbus_open() are -EXXX and vmbus_open() callers
are not supposed to analyze host error codes.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Implement the protocol for tearing down the monitor state established with
the host.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Tested-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
free_channel() has been invoked in
vmbus_remove() -> hv_process_channel_removal(), or vmbus_remove() ->
... -> vmbus_close_internal() -> hv_process_channel_removal().
We also change to use list_for_each_entry_safe(), because the entry
is removed in hv_process_channel_removal().
This patch fixes a bug in the vmbus unload path.
Thank Dan Carpenter for finding the issue!
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Cc: K. Y. Srinivasan <kys@microsoft.com>
Cc: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case we do request_resource() in vmbus_acpi_add() we need to tear it down
to be able to load the driver again. Otherwise the following crash in observed
when hv_vmbus unload/load sequence is performed on a Generation2 instance:
[ 38.165701] BUG: unable to handle kernel paging request at ffffffffa00075a0
[ 38.166315] IP: [<ffffffff8107dc5f>] __request_resource+0x2f/0x50
[ 38.166315] PGD 1f34067 PUD 1f35063 PMD 3f723067 PTE 0
[ 38.166315] Oops: 0000 [#1] SMP
[ 38.166315] Modules linked in: hv_vmbus(+) [last unloaded: hv_vmbus]
[ 38.166315] CPU: 0 PID: 267 Comm: modprobe Not tainted 3.19.0-rc5_bug923184+ #486
[ 38.166315] Hardware name: Microsoft Corporation Virtual Machine/Virtual Machine, BIOS Hyper-V UEFI Release v1.0 11/26/2012
[ 38.166315] task: ffff88003f401cb0 ti: ffff88003f60c000 task.ti: ffff88003f60c000
[ 38.166315] RIP: 0010:[<ffffffff8107dc5f>] [<ffffffff8107dc5f>] __request_resource+0x2f/0x50
[ 38.166315] RSP: 0018:ffff88003f60fb58 EFLAGS: 00010286
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Unify driver registration reporting and move it to debug level as normally daemons write to syslog themselves
and these kernel messages are useless.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Introduce FCOPY_VERSION_1 to support kernel replying to the negotiation
message with its own version.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Introduce VSS_OP_REGISTER1 to support kernel replying to the negotiation
message with its own version.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Convert to hv_utils_transport to support both netlink and /dev/vmbus/hv_kvp communication methods.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Unify the code with the recently introduced hv_utils_transport. Netlink
communication is disabled for fcopy.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Convert to hv_utils_transport to support both netlink and /dev/vmbus/hv_vss communication methods.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The intention is to make KVP/VSS drivers work through misc char devices.
Introduce an abstraction for kernel/userspace communication to make the
migration smoother. Transport operational mode (netlink or char device)
is determined by the first received message. To support driver upgrades
the switch from netlink to chardev operational mode is supported.
Every hv_util daemon is supposed to register 2 callbacks:
1) on_msg() to get notified when the userspace daemon sent a message;
2) on_reset() to get notified when the userspace daemon drops the connection.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Get an additional reference otherwise a crash is observed when hv_utils module is being unloaded while
fcopy daemon is still running. .owner gives us an additional reference when
someone holds a descriptor for the device.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Switch to using the hvutil_device_state state machine from using 3 different state variables:
fcopy_transaction.active, opened, and in_hand_shake.
State transitions are:
-> HVUTIL_DEVICE_INIT when driver loads or on device release
-> HVUTIL_READY if the handshake was successful
-> HVUTIL_HOSTMSG_RECEIVED when there is a non-negotiation message from the host
-> HVUTIL_USERSPACE_REQ after userspace daemon read the message
-> HVUTIL_USERSPACE_RECV after/if userspace has replied
-> HVUTIL_READY after we respond to the host
-> HVUTIL_DEVICE_DYING on driver unload
In hv_fcopy_onchannelcallback() process ICMSGTYPE_NEGOTIATE messages even when
the userspace daemon is disconnected, otherwise we can make the host think
we don't support FCOPY and disable the service completely.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Switch to using the hvutil_device_state state machine from using kvp_transaction.active.
State transitions are:
-> HVUTIL_DEVICE_INIT when driver loads or on device release
-> HVUTIL_READY if the handshake was successful
-> HVUTIL_HOSTMSG_RECEIVED when there is a non-negotiation message from the host
-> HVUTIL_USERSPACE_REQ after we sent the message to the userspace daemon
-> HVUTIL_USERSPACE_RECV after/if the userspace daemon has replied
-> HVUTIL_READY after we respond to the host
-> HVUTIL_DEVICE_DYING on driver unload
In hv_vss_onchannelcallback() process ICMSGTYPE_NEGOTIATE messages even when
the userspace daemon is disconnected, otherwise we can make the host think
we don't support VSS and disable the service completely.
Unfortunately there is no good way we can figure out that the userspace daemon
has died (unless we start treating all timeouts as such), add a protection
against processing new VSS_OP_REGISTER messages while being in the middle of a
transaction (HVUTIL_USERSPACE_REQ or HVUTIL_USERSPACE_RECV state).
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Switch to using the hvutil_device_state state machine from using 2 different state variables: kvp_transaction.active and
in_hand_shake.
State transitions are:
-> HVUTIL_DEVICE_INIT when driver loads or on device release
-> HVUTIL_READY if the handshake was successful
-> HVUTIL_HOSTMSG_RECEIVED when there is a non-negotiation message from the host
-> HVUTIL_USERSPACE_REQ after we sent the message to the userspace daemon
-> HVUTIL_USERSPACE_RECV after/if the userspace daemon has replied
-> HVUTIL_READY after we respond to the host
-> HVUTIL_DEVICE_DYING on driver unload
In hv_kvp_onchannelcallback() process ICMSGTYPE_NEGOTIATE messages even when
the userspace daemon is disconnected, otherwise we can make the host think
we don't support KVP and disable the service completely.
Unfortunately there is no good way we can figure out that the userspace daemon
has died (unless we start treating all timeouts as such). In case the daemon
restarts we skip the negotiation procedure (so the daemon is supposed to has
the same version). This behavior is unchanged from in_handshake approach.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
KVP/VSS/FCOPY drivers work in fully serialized mode: we wait till userspace
daemon registers, wait for a message from the host, send this message to the
daemon, get the reply, send it back to host, wait for another message.
Introduce enum hvutil_device_state to represend this state in all 3 drivers.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
'fcopy_work' (and fcopy_work_func) is a misnomer as it sounds like we expect
this useful work to happen and in reality it is just an emergency escape when
timeout happens.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
'kvp_work' (and kvp_work_func) is a misnomer as it sounds like we expect
this useful work to happen and in reality it is just an emergency escape when
timeout happens.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In theory, the host is not supposed to issue any requests before be reply to
the previous one. In KVP we, however, support the following scenarios:
1) A message was received before userspace daemon registered;
2) A message was received while the previous one is still being processed.
In VSS we support only the former. Add support for the later, use
hv_poll_channel() to do the job.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In theory, the host is not supposed to issue any requests before be reply to
the previous one. In KVP we, however, support the following scenarios:
1) A message was received before userspace daemon registered;
2) A message was received while the previous one is still being processed.
In FCOPY we support only the former. Add support for the later, use
hv_poll_channel() to do the job.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move poll_channel() to hyperv_vmbus.h and make it inline and rename it to hv_poll_channel() so it can be reused
in other hv_util modules.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
We set kvp_context when we want to postpone receiving a packet from vmbus due
to the previous transaction being unfinished. We, however, never reset this
state, all consequent kvp_respond_to_host() calls will result in poll_channel()
calling hv_kvp_onchannelcallback(). This doesn't cause real issues as:
1) Host is supposed to serialize transactions as well
2) If no message is pending vmbus_recvpacket() will return 0 recvlen.
This is just a cleanup.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
These declarations are internal to hv_util module and hv_fcopy_* declarations
already reside there.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Tested-by: Alex Ng <alexng@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
balloon_wrk.num_pages is __u32 and it comes from host in struct dm_balloon
where it is also __u32. We, however, use 'int' in balloon_up() and in case
we happen to receive num_pages>INT_MAX request we'll end up allocating zero
pages as 'num_pages < alloc_unit' check in alloc_balloon_pages() will pass.
Change num_pages type to unsigned int.
In real life ballooning request come with num_pages in [512, 32768] range so
this is more a future-proof/cleanup.
Reported-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
'Drivers: hv: hv_balloon: refuse to balloon below the floor' fix does not
correctly handle the case when val.freeram < num_pages as val.freeram is
__kernel_ulong_t and the 'val.freeram - num_pages' value will be a huge
positive value instead of being negative.
Usually host doesn't ask us to balloon more than val.freeram but in case
he have a memory hog started after we post the last pressure report we
can get into troubles.
Suggested-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Most of the retries can be done within a millisecond successfully, so we
sleep 1ms before the first retry, then gradually increase the retry
interval to 2^n with max value of 2048ms. Doing so, we will have shorter
overall delay time, because most of the cases succeed within 1-2 attempts.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Reviewed-by: K. Y. Srinivasan <kys@microsoft.com>
Reviewed-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
... and simplify alloc_balloon_pages() interface by removing redundant
alloc_error from it.
If we happen to enter balloon_up() with balloon_wrk.num_pages = 0 we will enter
infinite 'while (!done)' loop as alloc_balloon_pages() will be always returning
0 and not setting alloc_error. We will also be sending a meaningless message to
the host on every iteration.
The 'alloc_unit == 1 && alloc_error -> num_ballooned == 0' change and
alloc_error elimination requires a special comment. We do alloc_balloon_pages()
with 2 different alloc_unit values and there are 4 different
alloc_balloon_pages() results, let's check them all.
alloc_unit = 512:
1) num_ballooned = 0, alloc_error = 0: we do 'alloc_unit=1' and retry pre- and
post-patch.
2) num_ballooned > 0, alloc_error = 0: we check 'num_ballooned == num_pages'
and act accordingly, pre- and post-patch.
3) num_ballooned > 0, alloc_error > 0: we report this chunk and remain within
the loop, no changes here.
4) num_ballooned = 0, alloc_error > 0: we do 'alloc_unit=1' and retry pre- and
post-patch.
alloc_unit = 1:
1) num_ballooned = 0, alloc_error = 0: this can happen in two cases: when we
passed 'num_pages=0' to alloc_balloon_pages() or when there was no space in
bl_resp to place a single response. The second option is not possible as
bl_resp is of PAGE_SIZE size and single response 'union dm_mem_page_range' is
8 bytes, but the first one is (in theory, I think that Hyper-V host never
places such requests). Pre-patch code loops forever, post-patch code sends
a reply with more_pages = 0 and finishes.
2) num_ballooned > 0, alloc_error = 0: we ran out of space in bl_resp, we
report partial success and remain within the loop, no changes pre- and
post-patch.
3) num_ballooned > 0, alloc_error > 0: pre-patch code finishes, post-patch code
does one more try and if there is no progress (we finish with
'num_ballooned = 0') we finish. So we try a bit harder with this patch.
4) num_ballooned = 0, alloc_error > 0: both pre- and post-patch code enter
'more_pages = 0' branch and finish.
So this patch has two real effects:
1) We reply with an empty response to 'num_pages=0' request.
2) We try a bit harder on alloc_unit=1 allocations (and reply with an empty
tail reply in case we fail).
An empty reply should be supported by host as we were able to send it even with
pre-patch code when we were not able to allocate a single page.
Suggested-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 79208c57da ("Drivers: hv: hv_balloon: Make adjustments in computing
the floor") was inacurate as it introduced a jump in our piecewiese linear
'floor' function:
At 2048MB we have:
Left limit:
104 + 2048/8 = 360
Right limit:
256 + 2048/16 = 384 (so the right value is 232)
We now have to make an adjustment at 8192 boundary:
232 + 8192/16 = 744
512 + 8192/32 = 768 (so the right value is 488)
Suggested-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Currently we add memory in 128Mb blocks but the request from host can be
aligned differently. In such case we add a partially backed block and
when this block goes online we skip onlining pages which are not backed
(hv_online_page() callback serves this purpose). When we receive next
request for the same host add region we online pages which were not backed
before with hv_bring_pgs_online(). However, we don't check if the the block
in question was onlined and online this tail unconditionally. This is bad as
we avoid all online_pages() logic: these pages are not accounted, we don't
send notifications (and hv_balloon is not the only receiver of them),...
And, first of all, nobody asked as to online these pages. Solve the issue by
checking if the last previously backed page was onlined and onlining the tail
only in case it was.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It's not necessary any longer, since we can safely run the blocking
message handlers in vmbus_connection.work_queue now.
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Cc: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since the 2 fucntions can safely run in vmbus_connection.work_queue without
hang, we don't need to schedule new work items into the per-channel workqueue.
Actally we can even remove the per-channel workqueue now -- we'll do it
in the next patch.
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Cc: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A work item in vmbus_connection.work_queue can sleep, waiting for a new
host message (usually it is some kind of "completion" message). Currently
the new message will be handled in the same workqueue, but since work items
in the workqueue is serialized, we actually have no chance to handle
the new message if the current work item is sleeping -- as as result, the
current work item will hang forever.
K. Y. has posted the below fix to resolve the issue:
Drivers: hv: vmbus: Perform device register in the per-channel work element
Actually we can simplify the fix by directly running non-blocking message
handlers in the dispatch tasklet (inspired by K. Y.).
This patch is the fundamental change. The following 2 patches will simplify
the message offering and rescind-offering handling a lot.
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Cc: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Don't wait after sending request for offers to the host. This wait is
unnecessary and simply adds 5 seconds to the boot time.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Handle the case when the write to the ringbuffer fails. In this case,
unconditionally signal the host. Since we may have deferred signalling
the host based on the kick_q parameter, signalling the host
unconditionally in this case deals with the issue.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Export the vmbus_sendpacket_pagebuffer_ctl() interface. This export will be
used by the netvsc driver.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When a channel has been rescinded, the close operation is a noop.
Restructure the code so we deal with the rescind condition after
we properly cleanup the channel. I would like to thank
Dexuan Cui <decui@microsoft.com> for observing this problem.
The current code leaks memory when the channel is rescinded.
The current char-next branch is broken and this patch fixes
the bug.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The indenting makes it clear that there were curly braces intended here.
Fixes: 2dd37cb815 ('Drivers: hv: vmbus: Handle both rescind and offer messages in the same context')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
HV_CRASH_CTL_CRASH_NOTIFY is a 64 bit number. Depending on the usage context,
the value may be truncated. This patch is in response from the following
email from Wu Fengguang <fengguang.wu@intel.com>:
From: Wu Fengguang <fengguang.wu@intel.com>
Subject: [char-misc:char-misc-testing 25/45] drivers/hv/vmbus_drv.c:67:9: sparse:
constant 0x8000000000000000 is so big it is unsigned long
tree: git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc.git char-misc-testing
head: b3de8e3719
commit: 96c1d0581d [25/45] Drivers: hv: vmbus: Add support for VMBus panic notifier handler
reproduce:
# apt-get install sparse
git checkout 96c1d0581d
make ARCH=x86_64 allmodconfig
make C=1 CF=-D__CHECK_ENDIAN__
sparse warnings: (new ones prefixed by >>)
drivers/hv/vmbus_drv.c:67:9: sparse: constant 0x8000000000000000 is so big it is unsigned long
...
Signed-off-by: Nick Meier <nmeier@microsoft.com>
Reported-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Memory blocks can be onlined in random order. When this order is not natural
some memory pages are not onlined because of the redundant check in
hv_online_page().
Here is a real world scenario:
1) Host tries to hot-add the following (process_hot_add):
pg_start=rg_start=0x48000, pfn_cnt=111616, rg_size=262144
2) This results in adding 4 memory blocks:
[ 109.057866] init_memory_mapping: [mem 0x48000000-0x4fffffff]
[ 114.102698] init_memory_mapping: [mem 0x50000000-0x57ffffff]
[ 119.168039] init_memory_mapping: [mem 0x58000000-0x5fffffff]
[ 124.233053] init_memory_mapping: [mem 0x60000000-0x67ffffff]
The last one is incomplete but we have special has->covered_end_pfn counter to
avoid onlining non-backed frames and hv_bring_pgs_online() function to bring
them online later on.
3) Now we have 4 offline memory blocks: /sys/devices/system/memory/memory9-12
$ for f in /sys/devices/system/memory/memory*/state; do echo $f `cat $f`; done | grep -v onlin
/sys/devices/system/memory/memory10/state offline
/sys/devices/system/memory/memory11/state offline
/sys/devices/system/memory/memory12/state offline
/sys/devices/system/memory/memory9/state offline
4) We bring them online in non-natural order:
$grep MemTotal /proc/meminfo
MemTotal: 966348 kB
$echo online > /sys/devices/system/memory/memory12/state && grep MemTotal /proc/meminfo
MemTotal: 1019596 kB
$echo online > /sys/devices/system/memory/memory11/state && grep MemTotal /proc/meminfo
MemTotal: 1150668 kB
$echo online > /sys/devices/system/memory/memory9/state && grep MemTotal /proc/meminfo
MemTotal: 1150668 kB
As you can see memory9 block gives us zero additional memory. We can also
observe a huge discrepancy between host- and guest-reported memory sizes.
The root cause of the issue is the redundant pg >= covered_start_pfn check (and
covered_start_pfn advancing) in hv_online_page(). When upper memory block in
being onlined before the lower one (memory12 and memory11 in the above case) we
advance the covered_start_pfn pointer and all memory9 pages do not pass the
check. If the assumption that host always gives us requests in sequential order
and pg_start always equals rg_start when the first request for the new HA
region is received (that's the case in my testing) is correct than we can get
rid of covered_start_pfn and pg >= start_pfn check in hv_online_page() is
sufficient.
The current char-next branch is broken and this patch fixes
the bug.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
K. Y. Srinivasan
Vitaly Kuznetsov
Keith Mange
Dexuan Cui
Haiyang Zhang
Dan Carpenter
Nick Meier