* [PATCH V2 02/15] hyperv: Fix spelling of HV_UNKOWN
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 03/15] Drivers: hv: vmbus: Prevent sending data on a rescinded channel kys
` (13 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: Haiyang Zhang, K. Y. Srinivasan
From: Haiyang Zhang <haiyangz@microsoft.com>
Changed it to HV_UNKNOWN
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/channel_mgmt.c | 6 +++---
include/linux/hyperv.h | 2 +-
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
index dc6b675..f9c5827 100644
--- a/drivers/hv/channel_mgmt.c
+++ b/drivers/hv/channel_mgmt.c
@@ -134,7 +134,7 @@
},
/* Unknown GUID */
- { .dev_type = HV_UNKOWN,
+ { .dev_type = HV_UNKNOWN,
.perf_device = false,
},
};
@@ -163,9 +163,9 @@ static u16 hv_get_dev_type(const struct vmbus_channel *channel)
u16 i;
if (is_hvsock_channel(channel) || is_unsupported_vmbus_devs(guid))
- return HV_UNKOWN;
+ return HV_UNKNOWN;
- for (i = HV_IDE; i < HV_UNKOWN; i++) {
+ for (i = HV_IDE; i < HV_UNKNOWN; i++) {
if (!uuid_le_cmp(*guid, vmbus_devs[i].guid))
return i;
}
diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h
index 2a52d9a..35053f9 100644
--- a/include/linux/hyperv.h
+++ b/include/linux/hyperv.h
@@ -696,7 +696,7 @@ enum vmbus_device_type {
HV_FCOPY,
HV_BACKUP,
HV_DM,
- HV_UNKOWN,
+ HV_UNKNOWN,
};
struct vmbus_device {
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 03/15] Drivers: hv: vmbus: Prevent sending data on a rescinded channel
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
2016-12-03 20:34 ` [PATCH V2 02/15] hyperv: Fix spelling of HV_UNKOWN kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 04/15] Drivers: hv: vmbus: Enhance the rescind callback functionality kys
` (12 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: K. Y. Srinivasan <kys@microsoft.com>
After the channel is rescinded, the host does not read from the rescinded channel.
Fail writes to a channel that has already been rescinded.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/ring_buffer.c | 3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/drivers/hv/ring_buffer.c b/drivers/hv/ring_buffer.c
index cd49cb1..72f649d 100644
--- a/drivers/hv/ring_buffer.c
+++ b/drivers/hv/ring_buffer.c
@@ -298,6 +298,9 @@ int hv_ringbuffer_write(struct vmbus_channel *channel,
unsigned long flags = 0;
struct hv_ring_buffer_info *outring_info = &channel->outbound;
+ if (channel->rescind)
+ return -ENODEV;
+
for (i = 0; i < kv_count; i++)
totalbytes_towrite += kv_list[i].iov_len;
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 04/15] Drivers: hv: vmbus: Enhance the rescind callback functionality
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
2016-12-03 20:34 ` [PATCH V2 02/15] hyperv: Fix spelling of HV_UNKOWN kys
2016-12-03 20:34 ` [PATCH V2 03/15] Drivers: hv: vmbus: Prevent sending data on a rescinded channel kys
@ 2016-12-03 20:34 ` kys
2016-12-06 10:08 ` Greg KH
2016-12-03 20:34 ` [PATCH V2 05/15] hv: acquire vmbus_connection.channel_mutex in vmbus_free_channels() kys
` (11 subsequent siblings)
14 siblings, 1 reply; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: K. Y. Srinivasan <kys@microsoft.com>
Enhance the rescind callback functionality by permitting the passing of an opaque
pointer. This functionality will be used by vmbus device drivers to implement
rescind related cleanup more efficiently.
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/channel_mgmt.c | 11 +++++++----
include/linux/hyperv.h | 7 ++++---
2 files changed, 11 insertions(+), 7 deletions(-)
diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
index f9c5827..d83c1ac 100644
--- a/drivers/hv/channel_mgmt.c
+++ b/drivers/hv/channel_mgmt.c
@@ -825,8 +825,10 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr)
if (channel->device_obj) {
if (channel->chn_rescind_callback) {
- channel->chn_rescind_callback(channel);
- goto out;
+ channel->chn_rescind_callback(channel,
+ channel->rescind_arg);
+ if (is_hvsock_channel(channel))
+ goto out;
}
/*
* We will have to unregister this device from the
@@ -1215,9 +1217,10 @@ bool vmbus_are_subchannels_present(struct vmbus_channel *primary)
}
EXPORT_SYMBOL_GPL(vmbus_are_subchannels_present);
-void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel,
- void (*chn_rescind_cb)(struct vmbus_channel *))
+void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel, void *arg,
+ void (*chn_rescind_cb)(struct vmbus_channel *, void *))
{
channel->chn_rescind_callback = chn_rescind_cb;
+ channel->rescind_arg = arg;
}
EXPORT_SYMBOL_GPL(vmbus_set_chn_rescind_callback);
diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h
index 35053f9..a9a0ce4 100644
--- a/include/linux/hyperv.h
+++ b/include/linux/hyperv.h
@@ -800,7 +800,8 @@ struct vmbus_channel {
* Channel rescind callback. Some channels (the hvsock ones), need to
* register a callback which is invoked in vmbus_onoffer_rescind().
*/
- void (*chn_rescind_callback)(struct vmbus_channel *channel);
+ void (*chn_rescind_callback)(struct vmbus_channel *channel, void *arg);
+ void *rescind_arg;
/*
* The spinlock to protect the structure. It is being used to protect
@@ -957,8 +958,8 @@ static inline void clear_low_latency_mode(struct vmbus_channel *c)
void vmbus_set_sc_create_callback(struct vmbus_channel *primary_channel,
void (*sc_cr_cb)(struct vmbus_channel *new_sc));
-void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel,
- void (*chn_rescind_cb)(struct vmbus_channel *));
+void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel, void *arg,
+ void (*chn_rescind_cb)(struct vmbus_channel *, void *));
/*
* Retrieve the (sub) channel on which to send an outgoing request.
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH V2 04/15] Drivers: hv: vmbus: Enhance the rescind callback functionality
2016-12-03 20:34 ` [PATCH V2 04/15] Drivers: hv: vmbus: Enhance the rescind callback functionality kys
@ 2016-12-06 10:08 ` Greg KH
0 siblings, 0 replies; 23+ messages in thread
From: Greg KH @ 2016-12-06 10:08 UTC (permalink / raw)
To: kys; +Cc: linux-kernel, devel, olaf, apw, vkuznets, jasowang, leann.ogasawara
On Sat, Dec 03, 2016 at 12:34:31PM -0800, kys@exchange.microsoft.com wrote:
> From: K. Y. Srinivasan <kys@microsoft.com>
>
> Enhance the rescind callback functionality by permitting the passing of an opaque
> pointer. This functionality will be used by vmbus device drivers to implement
> rescind related cleanup more efficiently.
Eeek, why a void *? Please make this a _real_ pointer.
>
> Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
> ---
> drivers/hv/channel_mgmt.c | 11 +++++++----
> include/linux/hyperv.h | 7 ++++---
> 2 files changed, 11 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
> index f9c5827..d83c1ac 100644
> --- a/drivers/hv/channel_mgmt.c
> +++ b/drivers/hv/channel_mgmt.c
> @@ -825,8 +825,10 @@ static void vmbus_onoffer_rescind(struct vmbus_channel_message_header *hdr)
>
> if (channel->device_obj) {
> if (channel->chn_rescind_callback) {
> - channel->chn_rescind_callback(channel);
> - goto out;
> + channel->chn_rescind_callback(channel,
> + channel->rescind_arg);
> + if (is_hvsock_channel(channel))
> + goto out;
> }
> /*
> * We will have to unregister this device from the
> @@ -1215,9 +1217,10 @@ bool vmbus_are_subchannels_present(struct vmbus_channel *primary)
> }
> EXPORT_SYMBOL_GPL(vmbus_are_subchannels_present);
>
> -void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel,
> - void (*chn_rescind_cb)(struct vmbus_channel *))
> +void vmbus_set_chn_rescind_callback(struct vmbus_channel *channel, void *arg,
> + void (*chn_rescind_cb)(struct vmbus_channel *, void *))
> {
> channel->chn_rescind_callback = chn_rescind_cb;
> + channel->rescind_arg = arg;
> }
> EXPORT_SYMBOL_GPL(vmbus_set_chn_rescind_callback);
> diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h
> index 35053f9..a9a0ce4 100644
> --- a/include/linux/hyperv.h
> +++ b/include/linux/hyperv.h
> @@ -800,7 +800,8 @@ struct vmbus_channel {
> * Channel rescind callback. Some channels (the hvsock ones), need to
> * register a callback which is invoked in vmbus_onoffer_rescind().
> */
> - void (*chn_rescind_callback)(struct vmbus_channel *channel);
> + void (*chn_rescind_callback)(struct vmbus_channel *channel, void *arg);
> + void *rescind_arg;
Why does the channel have rescind_arg here?
Where do you use this functionality?
thanks,
greg k-h
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH V2 05/15] hv: acquire vmbus_connection.channel_mutex in vmbus_free_channels()
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (2 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 04/15] Drivers: hv: vmbus: Enhance the rescind callback functionality kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 06/15] hv: allocate synic pages for all present CPUs kys
` (10 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan, Stable
From: Vitaly Kuznetsov <vkuznets@redhat.com>
"kernel BUG at drivers/hv/channel_mgmt.c:350!" is observed when hv_vmbus
module is unloaded. BUG_ON() was introduced in commit 85d9aa705184
("Drivers: hv: vmbus: add an API vmbus_hvsock_device_unregister()") as
vmbus_free_channels() codepath was apparently forgotten.
Fixes: 85d9aa705184 ("Drivers: hv: vmbus: add an API vmbus_hvsock_device_unregister()")
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Cc: <Stable@vger.kernel.org>
---
drivers/hv/channel_mgmt.c | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)
diff --git a/drivers/hv/channel_mgmt.c b/drivers/hv/channel_mgmt.c
index d83c1ac..e68f06c 100644
--- a/drivers/hv/channel_mgmt.c
+++ b/drivers/hv/channel_mgmt.c
@@ -390,6 +390,7 @@ void vmbus_free_channels(void)
{
struct vmbus_channel *channel, *tmp;
+ mutex_lock(&vmbus_connection.channel_mutex);
list_for_each_entry_safe(channel, tmp, &vmbus_connection.chn_list,
listentry) {
/* hv_process_channel_removal() needs this */
@@ -397,6 +398,7 @@ void vmbus_free_channels(void)
vmbus_device_unregister(channel->device_obj);
}
+ mutex_unlock(&vmbus_connection.channel_mutex);
}
/*
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 06/15] hv: allocate synic pages for all present CPUs
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (3 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 05/15] hv: acquire vmbus_connection.channel_mutex in vmbus_free_channels() kys
@ 2016-12-03 20:34 ` kys
2016-12-06 10:09 ` Greg KH
2016-12-03 20:34 ` [PATCH V2 07/15] hv: init percpu_list in hv_synic_alloc() kys
` (9 subsequent siblings)
14 siblings, 1 reply; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: Vitaly Kuznetsov <vkuznets@redhat.com>
It may happen that not all CPUs are online when we do hv_synic_alloc() and
in case more CPUs come online later we may try accessing these allocated
structures.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/hv.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index 60dbd6c..e4bb498 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -411,7 +411,7 @@ int hv_synic_alloc(void)
goto err;
}
- for_each_online_cpu(cpu) {
+ for_each_present_cpu(cpu) {
hv_context.event_dpc[cpu] = kmalloc(size, GFP_ATOMIC);
if (hv_context.event_dpc[cpu] == NULL) {
pr_err("Unable to allocate event dpc\n");
@@ -482,7 +482,7 @@ void hv_synic_free(void)
int cpu;
kfree(hv_context.hv_numa_map);
- for_each_online_cpu(cpu)
+ for_each_present_cpu(cpu)
hv_synic_free_cpu(cpu);
}
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH V2 06/15] hv: allocate synic pages for all present CPUs
2016-12-03 20:34 ` [PATCH V2 06/15] hv: allocate synic pages for all present CPUs kys
@ 2016-12-06 10:09 ` Greg KH
0 siblings, 0 replies; 23+ messages in thread
From: Greg KH @ 2016-12-06 10:09 UTC (permalink / raw)
To: kys; +Cc: linux-kernel, devel, olaf, apw, vkuznets, jasowang, leann.ogasawara
On Sat, Dec 03, 2016 at 12:34:33PM -0800, kys@exchange.microsoft.com wrote:
> From: Vitaly Kuznetsov <vkuznets@redhat.com>
>
> It may happen that not all CPUs are online when we do hv_synic_alloc() and
> in case more CPUs come online later we may try accessing these allocated
> structures.
>
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
> ---
> drivers/hv/hv.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
For stable releases?
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH V2 07/15] hv: init percpu_list in hv_synic_alloc()
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (4 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 06/15] hv: allocate synic pages for all present CPUs kys
@ 2016-12-03 20:34 ` kys
2016-12-06 10:09 ` Greg KH
2016-12-03 20:34 ` [PATCH V2 08/15] hv: change clockevents unbind tactics kys
` (8 subsequent siblings)
14 siblings, 1 reply; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: Vitaly Kuznetsov <vkuznets@redhat.com>
Initializing hv_context.percpu_list in hv_synic_alloc() helps to prevent a
crash in percpu_channel_enq() when not all CPUs were online during
initialization and it naturally belongs there.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/hv.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index e4bb498..a2567a4 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -457,6 +457,8 @@ int hv_synic_alloc(void)
pr_err("Unable to allocate post msg page\n");
goto err;
}
+
+ INIT_LIST_HEAD(&hv_context.percpu_list[cpu]);
}
return 0;
@@ -552,8 +554,6 @@ void hv_synic_init(void *arg)
rdmsrl(HV_X64_MSR_VP_INDEX, vp_index);
hv_context.vp_index[cpu] = (u32)vp_index;
- INIT_LIST_HEAD(&hv_context.percpu_list[cpu]);
-
/*
* Register the per-cpu clockevent source.
*/
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH V2 07/15] hv: init percpu_list in hv_synic_alloc()
2016-12-03 20:34 ` [PATCH V2 07/15] hv: init percpu_list in hv_synic_alloc() kys
@ 2016-12-06 10:09 ` Greg KH
0 siblings, 0 replies; 23+ messages in thread
From: Greg KH @ 2016-12-06 10:09 UTC (permalink / raw)
To: kys; +Cc: linux-kernel, devel, olaf, apw, vkuznets, jasowang, leann.ogasawara
On Sat, Dec 03, 2016 at 12:34:34PM -0800, kys@exchange.microsoft.com wrote:
> From: Vitaly Kuznetsov <vkuznets@redhat.com>
>
> Initializing hv_context.percpu_list in hv_synic_alloc() helps to prevent a
> crash in percpu_channel_enq() when not all CPUs were online during
> initialization and it naturally belongs there.
>
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
> ---
> drivers/hv/hv.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
stable kernels?
^ permalink raw reply [flat|nested] 23+ messages in thread
* [PATCH V2 08/15] hv: change clockevents unbind tactics
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (5 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 07/15] hv: init percpu_list in hv_synic_alloc() kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 09/15] hv: switch to cpuhp state machine for synic init/cleanup kys
` (7 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: Vitaly Kuznetsov <vkuznets@redhat.com>
To get prepared to CPU offlining support we need co change the way how we
unbind clockevent devices. As one CPU may go online/offline multiple times
we need to bind it in hv_synic_init() and unbind it in hv_synic_cleanup().
There is an additional corner case: when we unload the module completely we
need to switch to some other clockevent mechanism before stopping VMBus or
we will hang. We can't call hv_synic_cleanup() before unloading VMBus as
we won't be able to send UNLOAD request and get a response so
hv_synic_clockevents_cleanup() has to live. Luckily, we can always call
clockevents_unbind_device(), even if it wasn't bound before and there is
no issue if we call it twice.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/hv.c | 6 ++++--
1 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index a2567a4..c11393c 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -575,7 +575,7 @@ void hv_synic_clockevents_cleanup(void)
if (!(ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE))
return;
- for_each_online_cpu(cpu)
+ for_each_present_cpu(cpu)
clockevents_unbind_device(hv_context.clk_evt[cpu], cpu);
}
@@ -594,8 +594,10 @@ void hv_synic_cleanup(void *arg)
return;
/* Turn off clockevent device */
- if (ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE)
+ if (ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE) {
+ clockevents_unbind_device(hv_context.clk_evt[cpu], cpu);
hv_ce_shutdown(hv_context.clk_evt[cpu]);
+ }
rdmsrl(HV_X64_MSR_SINT0 + VMBUS_MESSAGE_SINT, shared_sint.as_uint64);
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 09/15] hv: switch to cpuhp state machine for synic init/cleanup
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (6 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 08/15] hv: change clockevents unbind tactics kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 10/15] hv: make CPU offlining prevention fine-grained kys
` (6 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: Vitaly Kuznetsov <vkuznets@redhat.com>
To make it possible to online/offline CPUs switch to cpuhp infrastructure
for doing hv_synic_init()/hv_synic_cleanup().
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/hv.c | 15 +++++++--------
drivers/hv/hyperv_vmbus.h | 4 ++--
drivers/hv/vmbus_drv.c | 19 +++++++++++--------
3 files changed, 20 insertions(+), 18 deletions(-)
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index c11393c..b3f6a1b 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -495,7 +495,7 @@ void hv_synic_free(void)
* retrieve the initialized message and event pages. Otherwise, we create and
* initialize the message and event pages.
*/
-void hv_synic_init(void *arg)
+int hv_synic_init(unsigned int cpu)
{
u64 version;
union hv_synic_simp simp;
@@ -504,10 +504,8 @@ void hv_synic_init(void *arg)
union hv_synic_scontrol sctrl;
u64 vp_index;
- int cpu = smp_processor_id();
-
if (!hv_context.hypercall_page)
- return;
+ return -EFAULT;
/* Check the version */
rdmsrl(HV_X64_MSR_SVERSION, version);
@@ -562,7 +560,7 @@ void hv_synic_init(void *arg)
HV_TIMER_FREQUENCY,
HV_MIN_DELTA_TICKS,
HV_MAX_MAX_DELTA_TICKS);
- return;
+ return 0;
}
/*
@@ -582,16 +580,15 @@ void hv_synic_clockevents_cleanup(void)
/*
* hv_synic_cleanup - Cleanup routine for hv_synic_init().
*/
-void hv_synic_cleanup(void *arg)
+int hv_synic_cleanup(unsigned int cpu)
{
union hv_synic_sint shared_sint;
union hv_synic_simp simp;
union hv_synic_siefp siefp;
union hv_synic_scontrol sctrl;
- int cpu = smp_processor_id();
if (!hv_context.synic_initialized)
- return;
+ return -EFAULT;
/* Turn off clockevent device */
if (ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE) {
@@ -623,4 +620,6 @@ void hv_synic_cleanup(void *arg)
rdmsrl(HV_X64_MSR_SCONTROL, sctrl.as_uint64);
sctrl.enable = 0;
wrmsrl(HV_X64_MSR_SCONTROL, sctrl.as_uint64);
+
+ return 0;
}
diff --git a/drivers/hv/hyperv_vmbus.h b/drivers/hv/hyperv_vmbus.h
index 27982df..83beea7 100644
--- a/drivers/hv/hyperv_vmbus.h
+++ b/drivers/hv/hyperv_vmbus.h
@@ -505,9 +505,9 @@ extern int hv_post_message(union hv_connection_id connection_id,
extern void hv_synic_free(void);
-extern void hv_synic_init(void *irqarg);
+extern int hv_synic_init(unsigned int cpu);
-extern void hv_synic_cleanup(void *arg);
+extern int hv_synic_cleanup(unsigned int cpu);
extern void hv_synic_clockevents_cleanup(void);
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index 0276d2e..a115c90 100644
--- a/drivers/hv/vmbus_drv.c
+++ b/drivers/hv/vmbus_drv.c
@@ -49,6 +49,7 @@
static struct completion probe_event;
+static int hyperv_cpuhp_online;
static void hyperv_report_panic(struct pt_regs *regs)
{
@@ -844,7 +845,12 @@ static int vmbus_bus_init(void)
* Initialize the per-cpu interrupt state and
* connect to the host.
*/
- on_each_cpu(hv_synic_init, NULL, 1);
+ ret = cpuhp_setup_state(CPUHP_AP_ONLINE_DYN, "x86/hyperv:online",
+ hv_synic_init, hv_synic_cleanup);
+ if (ret < 0)
+ goto err_alloc;
+ hyperv_cpuhp_online = ret;
+
ret = vmbus_connect();
if (ret)
goto err_connect;
@@ -866,7 +872,7 @@ static int vmbus_bus_init(void)
return 0;
err_connect:
- on_each_cpu(hv_synic_cleanup, NULL, 1);
+ cpuhp_remove_state(hyperv_cpuhp_online);
err_alloc:
hv_synic_free();
hv_remove_vmbus_irq();
@@ -1320,12 +1326,9 @@ static int vmbus_acpi_add(struct acpi_device *device)
static void hv_kexec_handler(void)
{
- int cpu;
-
hv_synic_clockevents_cleanup();
vmbus_initiate_unload(false);
- for_each_online_cpu(cpu)
- smp_call_function_single(cpu, hv_synic_cleanup, NULL, 1);
+ cpuhp_remove_state(hyperv_cpuhp_online);
hv_cleanup(false);
};
@@ -1337,7 +1340,7 @@ static void hv_crash_handler(struct pt_regs *regs)
* doing the cleanup for current CPU only. This should be sufficient
* for kdump.
*/
- hv_synic_cleanup(NULL);
+ hv_synic_cleanup(smp_processor_id());
hv_cleanup(true);
};
@@ -1401,8 +1404,8 @@ static void __exit vmbus_exit(void)
hv_cleanup(false);
for_each_online_cpu(cpu) {
tasklet_kill(hv_context.event_dpc[cpu]);
- smp_call_function_single(cpu, hv_synic_cleanup, NULL, 1);
}
+ cpuhp_remove_state(hyperv_cpuhp_online);
hv_synic_free();
acpi_bus_unregister_driver(&vmbus_acpi_driver);
if (vmbus_proto_version > VERSION_WIN7)
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 10/15] hv: make CPU offlining prevention fine-grained
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (7 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 09/15] hv: switch to cpuhp state machine for synic init/cleanup kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 11/15] hv: don't reset hv_context.tsc_page on crash kys
` (5 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: Vitaly Kuznetsov <vkuznets@redhat.com>
Since commit e513229b4c38 ("Drivers: hv: vmbus: prevent cpu offlining on
newer hypervisors") cpu offlining was disabled. It is still true that we
can't offline CPUs which have VMBus channels bound to them but we may have
'free' CPUs (e.v. we booted with maxcpus= parameter and onlined CPUs after
VMBus was initialized), these CPUs may be disabled without issues.
In future, we may even allow closing CPUs which have only sub-channels
assinged to them by closing these sub-channels. All devices will continue
to work.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/hv.c | 31 +++++++++++++++++++++++++++++++
drivers/hv/vmbus_drv.c | 9 ++++-----
2 files changed, 35 insertions(+), 5 deletions(-)
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index b3f6a1b..4ece040 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -586,10 +586,41 @@ int hv_synic_cleanup(unsigned int cpu)
union hv_synic_simp simp;
union hv_synic_siefp siefp;
union hv_synic_scontrol sctrl;
+ struct vmbus_channel *channel, *sc;
+ bool channel_found = false;
+ unsigned long flags;
if (!hv_context.synic_initialized)
return -EFAULT;
+ /*
+ * Search for channels which are bound to the CPU we're about to
+ * cleanup. In case we find one and vmbus is still connected we need to
+ * fail, this will effectively prevent CPU offlining. There is no way
+ * we can re-bind channels to different CPUs for now.
+ */
+ mutex_lock(&vmbus_connection.channel_mutex);
+ list_for_each_entry(channel, &vmbus_connection.chn_list, listentry) {
+ if (channel->target_cpu == cpu) {
+ channel_found = true;
+ break;
+ }
+ spin_lock_irqsave(&channel->lock, flags);
+ list_for_each_entry(sc, &channel->sc_list, sc_list) {
+ if (sc->target_cpu == cpu) {
+ channel_found = true;
+ break;
+ }
+ }
+ spin_unlock_irqrestore(&channel->lock, flags);
+ if (channel_found)
+ break;
+ }
+ mutex_unlock(&vmbus_connection.channel_mutex);
+
+ if (channel_found && vmbus_connection.conn_state == CONNECTED)
+ return -EBUSY;
+
/* Turn off clockevent device */
if (ms_hyperv.features & HV_X64_MSR_SYNTIMER_AVAILABLE) {
clockevents_unbind_device(hv_context.clk_evt[cpu], cpu);
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index a115c90..61d879f 100644
--- a/drivers/hv/vmbus_drv.c
+++ b/drivers/hv/vmbus_drv.c
@@ -855,9 +855,6 @@ static int vmbus_bus_init(void)
if (ret)
goto err_connect;
- if (vmbus_proto_version > VERSION_WIN7)
- cpu_hotplug_disable();
-
/*
* Only register if the crash MSRs are available
*/
@@ -1328,6 +1325,9 @@ static void hv_kexec_handler(void)
{
hv_synic_clockevents_cleanup();
vmbus_initiate_unload(false);
+ vmbus_connection.conn_state = DISCONNECTED;
+ /* Make sure conn_state is set as hv_synic_cleanup checks for it */
+ mb();
cpuhp_remove_state(hyperv_cpuhp_online);
hv_cleanup(false);
};
@@ -1340,6 +1340,7 @@ static void hv_crash_handler(struct pt_regs *regs)
* doing the cleanup for current CPU only. This should be sufficient
* for kdump.
*/
+ vmbus_connection.conn_state = DISCONNECTED;
hv_synic_cleanup(smp_processor_id());
hv_cleanup(true);
};
@@ -1408,8 +1409,6 @@ static void __exit vmbus_exit(void)
cpuhp_remove_state(hyperv_cpuhp_online);
hv_synic_free();
acpi_bus_unregister_driver(&vmbus_acpi_driver);
- if (vmbus_proto_version > VERSION_WIN7)
- cpu_hotplug_enable();
}
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 11/15] hv: don't reset hv_context.tsc_page on crash
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (8 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 10/15] hv: make CPU offlining prevention fine-grained kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 12/15] vmbus: add support for dynamic device id's kys
` (4 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: K. Y. Srinivasan
From: Vitaly Kuznetsov <vkuznets@redhat.com>
It may happen that secondary CPUs are still alive and resetting
hv_context.tsc_page will cause a consequent crash in read_hv_clock_tsc()
as we don't check for it being not NULL there. It is safe as we're not
freeing this page anyways.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/hv.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/drivers/hv/hv.c b/drivers/hv/hv.c
index 4ece040..04a0fd8 100644
--- a/drivers/hv/hv.c
+++ b/drivers/hv/hv.c
@@ -309,9 +309,10 @@ void hv_cleanup(bool crash)
hypercall_msr.as_uint64 = 0;
wrmsrl(HV_X64_MSR_REFERENCE_TSC, hypercall_msr.as_uint64);
- if (!crash)
+ if (!crash) {
vfree(hv_context.tsc_page);
- hv_context.tsc_page = NULL;
+ hv_context.tsc_page = NULL;
+ }
}
#endif
}
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 12/15] vmbus: add support for dynamic device id's
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (9 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 11/15] hv: don't reset hv_context.tsc_page on crash kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 13/15] uio-hv-generic: new userspace i/o driver for VMBus kys
` (3 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: Stephen Hemminger, K. Y. Srinivasan
From: Stephen Hemminger <sthemmin@microsoft.com>
This patch adds sysfs interface to dynamically bind new UUID values
to existing VMBus device. This is useful for generic UIO driver to
act similar to uio_pci_generic.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/vmbus_drv.c | 174 +++++++++++++++++++++++++++++++++++++++++++++--
include/linux/hyperv.h | 6 ++
2 files changed, 172 insertions(+), 8 deletions(-)
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index 61d879f..565bdd1 100644
--- a/drivers/hv/vmbus_drv.c
+++ b/drivers/hv/vmbus_drv.c
@@ -45,6 +45,11 @@
#include <linux/random.h>
#include "hyperv_vmbus.h"
+struct vmbus_dynid {
+ struct list_head node;
+ struct hv_vmbus_device_id id;
+};
+
static struct acpi_device *hv_acpi_dev;
static struct completion probe_event;
@@ -501,7 +506,7 @@ static ssize_t device_show(struct device *dev,
static DEVICE_ATTR_RO(device);
/* Set up per device attributes in /sys/bus/vmbus/devices/<bus device> */
-static struct attribute *vmbus_attrs[] = {
+static struct attribute *vmbus_dev_attrs[] = {
&dev_attr_id.attr,
&dev_attr_state.attr,
&dev_attr_monitor_id.attr,
@@ -529,7 +534,7 @@ static ssize_t device_show(struct device *dev,
&dev_attr_device.attr,
NULL,
};
-ATTRIBUTE_GROUPS(vmbus);
+ATTRIBUTE_GROUPS(vmbus_dev);
/*
* vmbus_uevent - add uevent for our device
@@ -566,10 +571,29 @@ static inline bool is_null_guid(const uuid_le *guid)
* Return a matching hv_vmbus_device_id pointer.
* If there is no match, return NULL.
*/
-static const struct hv_vmbus_device_id *hv_vmbus_get_id(
- const struct hv_vmbus_device_id *id,
+static const struct hv_vmbus_device_id *hv_vmbus_get_id(struct hv_driver *drv,
const uuid_le *guid)
{
+ const struct hv_vmbus_device_id *id = NULL;
+ struct vmbus_dynid *dynid;
+
+ /* Look at the dynamic ids first, before the static ones */
+ spin_lock(&drv->dynids.lock);
+ list_for_each_entry(dynid, &drv->dynids.list, node) {
+ if (!uuid_le_cmp(dynid->id.guid, *guid)) {
+ id = &dynid->id;
+ break;
+ }
+ }
+ spin_unlock(&drv->dynids.lock);
+
+ if (id)
+ return id;
+
+ id = drv->id_table;
+ if (id == NULL)
+ return NULL; /* empty device table */
+
for (; !is_null_guid(&id->guid); id++)
if (!uuid_le_cmp(id->guid, *guid))
return id;
@@ -577,6 +601,134 @@ static inline bool is_null_guid(const uuid_le *guid)
return NULL;
}
+/* vmbus_add_dynid - add a new device ID to this driver and re-probe devices */
+static int vmbus_add_dynid(struct hv_driver *drv, uuid_le *guid)
+{
+ struct vmbus_dynid *dynid;
+
+ dynid = kzalloc(sizeof(*dynid), GFP_KERNEL);
+ if (!dynid)
+ return -ENOMEM;
+
+ dynid->id.guid = *guid;
+
+ spin_lock(&drv->dynids.lock);
+ list_add_tail(&dynid->node, &drv->dynids.list);
+ spin_unlock(&drv->dynids.lock);
+
+ return driver_attach(&drv->driver);
+}
+
+static void vmbus_free_dynids(struct hv_driver *drv)
+{
+ struct vmbus_dynid *dynid, *n;
+
+ spin_lock(&drv->dynids.lock);
+ list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) {
+ list_del(&dynid->node);
+ kfree(dynid);
+ }
+ spin_unlock(&drv->dynids.lock);
+}
+
+/* Parse string of form: 1b4e28ba-2fa1-11d2-883f-b9a761bde3f */
+static int get_uuid_le(const char *str, uuid_le *uu)
+{
+ unsigned int b[16];
+ int i;
+
+ if (strlen(str) < 37)
+ return -1;
+
+ for (i = 0; i < 36; i++) {
+ switch (i) {
+ case 8: case 13: case 18: case 23:
+ if (str[i] != '-')
+ return -1;
+ break;
+ default:
+ if (!isxdigit(str[i]))
+ return -1;
+ }
+ }
+
+ /* unparse little endian output byte order */
+ if (sscanf(str,
+ "%2x%2x%2x%2x-%2x%2x-%2x%2x-%2x%2x-%2x%2x%2x%2x%2x%2x",
+ &b[3], &b[2], &b[1], &b[0],
+ &b[5], &b[4], &b[7], &b[6], &b[8], &b[9],
+ &b[10], &b[11], &b[12], &b[13], &b[14], &b[15]) != 16)
+ return -1;
+
+ for (i = 0; i < 16; i++)
+ uu->b[i] = b[i];
+ return 0;
+}
+
+/*
+ * store_new_id - sysfs frontend to vmbus_add_dynid()
+ *
+ * Allow GUIDs to be added to an existing driver via sysfs.
+ */
+static ssize_t new_id_store(struct device_driver *driver, const char *buf,
+ size_t count)
+{
+ struct hv_driver *drv = drv_to_hv_drv(driver);
+ uuid_le guid = NULL_UUID_LE;
+ ssize_t retval;
+
+ if (get_uuid_le(buf, &guid) != 0)
+ return -EINVAL;
+
+ if (hv_vmbus_get_id(drv, &guid))
+ return -EEXIST;
+
+ retval = vmbus_add_dynid(drv, &guid);
+ if (retval)
+ return retval;
+ return count;
+}
+static DRIVER_ATTR_WO(new_id);
+
+/*
+ * store_remove_id - remove a PCI device ID from this driver
+ *
+ * Removes a dynamic pci device ID to this driver.
+ */
+static ssize_t remove_id_store(struct device_driver *driver, const char *buf,
+ size_t count)
+{
+ struct hv_driver *drv = drv_to_hv_drv(driver);
+ struct vmbus_dynid *dynid, *n;
+ uuid_le guid = NULL_UUID_LE;
+ size_t retval = -ENODEV;
+
+ if (get_uuid_le(buf, &guid))
+ return -EINVAL;
+
+ spin_lock(&drv->dynids.lock);
+ list_for_each_entry_safe(dynid, n, &drv->dynids.list, node) {
+ struct hv_vmbus_device_id *id = &dynid->id;
+
+ if (!uuid_le_cmp(id->guid, guid)) {
+ list_del(&dynid->node);
+ kfree(dynid);
+ retval = count;
+ break;
+ }
+ }
+ spin_unlock(&drv->dynids.lock);
+
+ return retval;
+}
+static DRIVER_ATTR_WO(remove_id);
+
+static struct attribute *vmbus_drv_attrs[] = {
+ &driver_attr_new_id.attr,
+ &driver_attr_remove_id.attr,
+ NULL,
+};
+ATTRIBUTE_GROUPS(vmbus_drv);
/*
@@ -591,7 +743,7 @@ static int vmbus_match(struct device *device, struct device_driver *driver)
if (is_hvsock_channel(hv_dev->channel))
return drv->hvsock;
- if (hv_vmbus_get_id(drv->id_table, &hv_dev->dev_type))
+ if (hv_vmbus_get_id(drv, &hv_dev->dev_type))
return 1;
return 0;
@@ -608,7 +760,7 @@ static int vmbus_probe(struct device *child_device)
struct hv_device *dev = device_to_hv_device(child_device);
const struct hv_vmbus_device_id *dev_id;
- dev_id = hv_vmbus_get_id(drv->id_table, &dev->dev_type);
+ dev_id = hv_vmbus_get_id(drv, &dev->dev_type);
if (drv->probe) {
ret = drv->probe(dev, dev_id);
if (ret != 0)
@@ -685,7 +837,8 @@ static void vmbus_device_release(struct device *device)
.remove = vmbus_remove,
.probe = vmbus_probe,
.uevent = vmbus_uevent,
- .dev_groups = vmbus_groups,
+ .dev_groups = vmbus_dev_groups,
+ .drv_groups = vmbus_drv_groups,
};
struct onmessage_work_context {
@@ -908,6 +1061,9 @@ int __vmbus_driver_register(struct hv_driver *hv_driver, struct module *owner, c
hv_driver->driver.mod_name = mod_name;
hv_driver->driver.bus = &hv_bus;
+ spin_lock_init(&hv_driver->dynids.lock);
+ INIT_LIST_HEAD(&hv_driver->dynids.list);
+
ret = driver_register(&hv_driver->driver);
return ret;
@@ -926,8 +1082,10 @@ void vmbus_driver_unregister(struct hv_driver *hv_driver)
{
pr_info("unregistering driver %s\n", hv_driver->name);
- if (!vmbus_exists())
+ if (!vmbus_exists()) {
driver_unregister(&hv_driver->driver);
+ vmbus_free_dynids(hv_driver);
+ }
}
EXPORT_SYMBOL_GPL(vmbus_driver_unregister);
diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h
index a9a0ce4..1d0ebe9 100644
--- a/include/linux/hyperv.h
+++ b/include/linux/hyperv.h
@@ -1120,6 +1120,12 @@ struct hv_driver {
struct device_driver driver;
+ /* dynamic device GUID's */
+ struct {
+ spinlock_t lock;
+ struct list_head list;
+ } dynids;
+
int (*probe)(struct hv_device *, const struct hv_vmbus_device_id *);
int (*remove)(struct hv_device *);
void (*shutdown)(struct hv_device *);
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 13/15] uio-hv-generic: new userspace i/o driver for VMBus
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (10 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 12/15] vmbus: add support for dynamic device id's kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 14/15] Tools: hv: kvp: configurable external scripts path kys
` (2 subsequent siblings)
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: Stephen Hemminger, K. Y. Srinivasan
From: Stephen Hemminger <sthemmin@microsoft.com>
This is a new driver to enable userspace networking on VMBus.
It is based largely on the similar driver that already exists
for PCI, and earlier work done by Brocade to support DPDK.
Signed-off-by: Stephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
MAINTAINERS | 1 +
drivers/hv/connection.c | 1 +
drivers/uio/Kconfig | 9 ++
drivers/uio/Makefile | 1 +
drivers/uio/uio_hv_generic.c | 218 ++++++++++++++++++++++++++++++++++++++++++
5 files changed, 230 insertions(+), 0 deletions(-)
create mode 100644 drivers/uio/uio_hv_generic.c
diff --git a/MAINTAINERS b/MAINTAINERS
index 2f89c3b..6d6bea9 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -5851,6 +5851,7 @@ F: drivers/input/serio/hyperv-keyboard.c
F: drivers/pci/host/pci-hyperv.c
F: drivers/net/hyperv/
F: drivers/scsi/storvsc_drv.c
+F: drivers/uio/uio_hv_generic.c
F: drivers/video/fbdev/hyperv_fb.c
F: include/linux/hyperv.h
F: tools/hv/
diff --git a/drivers/hv/connection.c b/drivers/hv/connection.c
index 840b6db..9b72ebc 100644
--- a/drivers/hv/connection.c
+++ b/drivers/hv/connection.c
@@ -39,6 +39,7 @@ struct vmbus_connection vmbus_connection = {
.conn_state = DISCONNECTED,
.next_gpadl_handle = ATOMIC_INIT(0xE1E10),
};
+EXPORT_SYMBOL_GPL(vmbus_connection);
/*
* Negotiated protocol version with the host.
diff --git a/drivers/uio/Kconfig b/drivers/uio/Kconfig
index 52c98ce..7e8dc78 100644
--- a/drivers/uio/Kconfig
+++ b/drivers/uio/Kconfig
@@ -155,4 +155,13 @@ config UIO_MF624
If you compile this as a module, it will be called uio_mf624.
+config UIO_HV_GENERIC
+ tristate "Generic driver for Hyper-V VMBus"
+ depends on HYPERV
+ help
+ Generic driver that you can bind, dynamically, to any
+ Hyper-V VMBus device. It is useful to provide direct access
+ to network and storage devices from userspace.
+
+ If you compile this as a module, it will be called uio_hv_generic.
endif
diff --git a/drivers/uio/Makefile b/drivers/uio/Makefile
index 8560dad..e9663bb 100644
--- a/drivers/uio/Makefile
+++ b/drivers/uio/Makefile
@@ -9,3 +9,4 @@ obj-$(CONFIG_UIO_NETX) += uio_netx.o
obj-$(CONFIG_UIO_PRUSS) += uio_pruss.o
obj-$(CONFIG_UIO_MF624) += uio_mf624.o
obj-$(CONFIG_UIO_FSL_ELBC_GPCM) += uio_fsl_elbc_gpcm.o
+obj-$(CONFIG_UIO_HV_GENERIC) += uio_hv_generic.o
diff --git a/drivers/uio/uio_hv_generic.c b/drivers/uio/uio_hv_generic.c
new file mode 100644
index 0000000..ad3ab58
--- /dev/null
+++ b/drivers/uio/uio_hv_generic.c
@@ -0,0 +1,218 @@
+/*
+ * uio_hv_generic - generic UIO driver for VMBus
+ *
+ * Copyright (c) 2013-2016 Brocade Communications Systems, Inc.
+ * Copyright (c) 2016, Microsoft Corporation.
+ *
+ *
+ * This work is licensed under the terms of the GNU GPL, version 2.
+ *
+ * Since the driver does not declare any device ids, you must allocate
+ * id and bind the device to the driver yourself. For example:
+ *
+ * # echo "f8615163-df3e-46c5-913f-f2d2f965ed0e" \
+ * > /sys/bus/vmbus/drivers/uio_hv_generic
+ * # echo -n vmbus-ed963694-e847-4b2a-85af-bc9cfc11d6f3 \
+ * > /sys/bus/vmbus/drivers/hv_netvsc/unbind
+ * # echo -n vmbus-ed963694-e847-4b2a-85af-bc9cfc11d6f3 \
+ * > /sys/bus/vmbus/drivers/uio_hv_generic/bind
+ */
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
+#include <linux/device.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/uio_driver.h>
+#include <linux/netdevice.h>
+#include <linux/if_ether.h>
+#include <linux/skbuff.h>
+#include <linux/hyperv.h>
+#include <linux/vmalloc.h>
+#include <linux/slab.h>
+
+#include "../hv/hyperv_vmbus.h"
+
+#define DRIVER_VERSION "0.02.0"
+#define DRIVER_AUTHOR "Stephen Hemminger <sthemmin at microsoft.com>"
+#define DRIVER_DESC "Generic UIO driver for VMBus devices"
+
+/*
+ * List of resources to be mapped to user space
+ * can be extended up to MAX_UIO_MAPS(5) items
+ */
+enum hv_uio_map {
+ TXRX_RING_MAP = 0,
+ INT_PAGE_MAP,
+ MON_PAGE_MAP,
+};
+
+#define HV_RING_SIZE 512
+
+struct hv_uio_private_data {
+ struct uio_info info;
+ struct hv_device *device;
+};
+
+static int
+hv_uio_mmap(struct uio_info *info, struct vm_area_struct *vma)
+{
+ int mi;
+
+ if (vma->vm_pgoff >= MAX_UIO_MAPS)
+ return -EINVAL;
+
+ if (info->mem[vma->vm_pgoff].size == 0)
+ return -EINVAL;
+
+ mi = (int)vma->vm_pgoff;
+
+ return remap_pfn_range(vma, vma->vm_start,
+ virt_to_phys((void *)info->mem[mi].addr) >> PAGE_SHIFT,
+ vma->vm_end - vma->vm_start, vma->vm_page_prot);
+}
+
+/*
+ * This is the irqcontrol callback to be registered to uio_info.
+ * It can be used to disable/enable interrupt from user space processes.
+ *
+ * @param info
+ * pointer to uio_info.
+ * @param irq_state
+ * state value. 1 to enable interrupt, 0 to disable interrupt.
+ */
+static int
+hv_uio_irqcontrol(struct uio_info *info, s32 irq_state)
+{
+ struct hv_uio_private_data *pdata = info->priv;
+ struct hv_device *dev = pdata->device;
+
+ dev->channel->inbound.ring_buffer->interrupt_mask = !irq_state;
+ virt_mb();
+
+ return 0;
+}
+
+/*
+ * Callback from vmbus_event when something is in inbound ring.
+ */
+static void hv_uio_channel_cb(void *context)
+{
+ struct hv_uio_private_data *pdata = context;
+ struct hv_device *dev = pdata->device;
+
+ dev->channel->inbound.ring_buffer->interrupt_mask = 1;
+ virt_mb();
+
+ uio_event_notify(&pdata->info);
+}
+
+static int
+hv_uio_probe(struct hv_device *dev,
+ const struct hv_vmbus_device_id *dev_id)
+{
+ struct hv_uio_private_data *pdata;
+ int ret;
+
+ pdata = kzalloc(sizeof(*pdata), GFP_KERNEL);
+ if (!pdata)
+ return -ENOMEM;
+
+ ret = vmbus_open(dev->channel, HV_RING_SIZE * PAGE_SIZE,
+ HV_RING_SIZE * PAGE_SIZE, NULL, 0,
+ hv_uio_channel_cb, pdata);
+ if (ret)
+ goto fail;
+
+ dev->channel->inbound.ring_buffer->interrupt_mask = 1;
+ dev->channel->batched_reading = false;
+
+ /* Fill general uio info */
+ pdata->info.name = "uio_hv_generic";
+ pdata->info.version = DRIVER_VERSION;
+ pdata->info.irqcontrol = hv_uio_irqcontrol;
+ pdata->info.mmap = hv_uio_mmap;
+ pdata->info.irq = UIO_IRQ_CUSTOM;
+
+ /* mem resources */
+ pdata->info.mem[TXRX_RING_MAP].name = "txrx_rings";
+ pdata->info.mem[TXRX_RING_MAP].addr
+ = (phys_addr_t)dev->channel->ringbuffer_pages;
+ pdata->info.mem[TXRX_RING_MAP].size
+ = dev->channel->ringbuffer_pagecount * PAGE_SIZE;
+ pdata->info.mem[TXRX_RING_MAP].memtype = UIO_MEM_LOGICAL;
+
+ pdata->info.mem[INT_PAGE_MAP].name = "int_page";
+ pdata->info.mem[INT_PAGE_MAP].addr =
+ (phys_addr_t)vmbus_connection.int_page;
+ pdata->info.mem[INT_PAGE_MAP].size = PAGE_SIZE;
+ pdata->info.mem[INT_PAGE_MAP].memtype = UIO_MEM_LOGICAL;
+
+ pdata->info.mem[MON_PAGE_MAP].name = "monitor_pages";
+ pdata->info.mem[MON_PAGE_MAP].addr =
+ (phys_addr_t)vmbus_connection.monitor_pages[1];
+ pdata->info.mem[MON_PAGE_MAP].size = PAGE_SIZE;
+ pdata->info.mem[MON_PAGE_MAP].memtype = UIO_MEM_LOGICAL;
+
+ pdata->info.priv = pdata;
+ pdata->device = dev;
+
+ ret = uio_register_device(&dev->device, &pdata->info);
+ if (ret) {
+ dev_err(&dev->device, "hv_uio register failed\n");
+ goto fail_close;
+ }
+
+ hv_set_drvdata(dev, pdata);
+
+ return 0;
+
+fail_close:
+ vmbus_close(dev->channel);
+fail:
+ kfree(pdata);
+
+ return ret;
+}
+
+static int
+hv_uio_remove(struct hv_device *dev)
+{
+ struct hv_uio_private_data *pdata = hv_get_drvdata(dev);
+
+ if (!pdata)
+ return 0;
+
+ uio_unregister_device(&pdata->info);
+ hv_set_drvdata(dev, NULL);
+ vmbus_close(dev->channel);
+ kfree(pdata);
+ return 0;
+}
+
+static struct hv_driver hv_uio_drv = {
+ .name = "uio_hv_generic",
+ .id_table = NULL, /* only dynamic id's */
+ .probe = hv_uio_probe,
+ .remove = hv_uio_remove,
+};
+
+static int __init
+hyperv_module_init(void)
+{
+ return vmbus_driver_register(&hv_uio_drv);
+}
+
+static void __exit
+hyperv_module_exit(void)
+{
+ vmbus_driver_unregister(&hv_uio_drv);
+}
+
+module_init(hyperv_module_init);
+module_exit(hyperv_module_exit);
+
+MODULE_VERSION(DRIVER_VERSION);
+MODULE_LICENSE("GPL v2");
+MODULE_AUTHOR(DRIVER_AUTHOR);
+MODULE_DESCRIPTION(DRIVER_DESC);
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 14/15] Tools: hv: kvp: configurable external scripts path
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (11 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 13/15] uio-hv-generic: new userspace i/o driver for VMBus kys
@ 2016-12-03 20:34 ` kys
2016-12-03 20:34 ` [PATCH V2 15/15] hyperv: Add a function to detect if the device is a vmbus dev kys
2016-12-06 10:07 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() Greg KH
14 siblings, 0 replies; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: Alex Fluter, K. Y. Srinivasan
From: Alex Fluter <afluter@yandex.com>
error when running hypervkvpd:
$ sudo ./hv_kvp_daemon -n
sh: hv_get_dns_info: command not found
sh: hv_get_dhcp_info: command not found
sh: hv_get_dns_info: command not found
sh: hv_get_dhcp_info: command not found
The external scripts are not installed in system path,
adding a configurable macro.
Signed-off-by: Alex Fluter <afluter@yandex.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
tools/hv/hv_kvp_daemon.c | 11 ++++++++---
1 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/tools/hv/hv_kvp_daemon.c b/tools/hv/hv_kvp_daemon.c
index d791dbf..f1758fc 100644
--- a/tools/hv/hv_kvp_daemon.c
+++ b/tools/hv/hv_kvp_daemon.c
@@ -93,6 +93,10 @@ enum {
#define KVP_CONFIG_LOC "/var/lib/hyperv"
+#ifndef KVP_SCRIPTS_PATH
+#define KVP_SCRIPTS_PATH "/usr/libexec/hypervkvpd/"
+#endif
+
#define MAX_FILE_NAME 100
#define ENTRIES_PER_BLOCK 50
@@ -818,7 +822,7 @@ static void kvp_get_ipconfig_info(char *if_name,
* .
*/
- sprintf(cmd, "%s", "hv_get_dns_info");
+ sprintf(cmd, KVP_SCRIPTS_PATH "%s", "hv_get_dns_info");
/*
* Execute the command to gather DNS info.
@@ -835,7 +839,7 @@ static void kvp_get_ipconfig_info(char *if_name,
* Enabled: DHCP enabled.
*/
- sprintf(cmd, "%s %s", "hv_get_dhcp_info", if_name);
+ sprintf(cmd, KVP_SCRIPTS_PATH "%s %s", "hv_get_dhcp_info", if_name);
file = popen(cmd, "r");
if (file == NULL)
@@ -1341,7 +1345,8 @@ static int kvp_set_ip_info(char *if_name, struct hv_kvp_ipaddr_value *new_val)
* invoke the external script to do its magic.
*/
- snprintf(cmd, sizeof(cmd), "%s %s", "hv_set_ifconfig", if_file);
+ snprintf(cmd, sizeof(cmd), KVP_SCRIPTS_PATH "%s %s",
+ "hv_set_ifconfig", if_file);
if (system(cmd)) {
syslog(LOG_ERR, "Failed to execute cmd '%s'; error: %d %s",
cmd, errno, strerror(errno));
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* [PATCH V2 15/15] hyperv: Add a function to detect if the device is a vmbus dev
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (12 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 14/15] Tools: hv: kvp: configurable external scripts path kys
@ 2016-12-03 20:34 ` kys
2016-12-06 10:54 ` Greg KH
2016-12-06 10:07 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() Greg KH
14 siblings, 1 reply; 23+ messages in thread
From: kys @ 2016-12-03 20:34 UTC (permalink / raw)
To: gregkh, linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara
Cc: Haiyang Zhang, K. Y. Srinivasan
From: Haiyang Zhang <haiyangz@microsoft.com>
On Hyper-V, every VF interface has a corresponding synthetic
interface managed by netvsc that share the same MAC
address. netvsc registers for netdev events to manage this
association. Currently we use the MAC address to manage this
association but going forward, we want to use a serial number that
the host publishes. To do this we need functionality similar
to dev_is_pci() for the vmbus devices. Implement such a function.
This function will be used in subsequent patches to netvsc and in the
interest of eliminating cross tree dependencies, this patch is being
submitted first.
Signed-off-by: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
---
drivers/hv/vmbus_drv.c | 6 ++++++
include/linux/hyperv.h | 2 ++
2 files changed, 8 insertions(+), 0 deletions(-)
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index 565bdd1..fe96f1e 100644
--- a/drivers/hv/vmbus_drv.c
+++ b/drivers/hv/vmbus_drv.c
@@ -846,6 +846,12 @@ struct onmessage_work_context {
struct hv_message msg;
};
+bool dev_is_vmbus(struct device *dev)
+{
+ return dev->bus == &hv_bus;
+}
+EXPORT_SYMBOL_GPL(dev_is_vmbus);
+
static void vmbus_onmessage_work(struct work_struct *work)
{
struct onmessage_work_context *ctx;
diff --git a/include/linux/hyperv.h b/include/linux/hyperv.h
index 1d0ebe9..918c15b 100644
--- a/include/linux/hyperv.h
+++ b/include/linux/hyperv.h
@@ -947,6 +947,8 @@ static inline void clear_low_latency_mode(struct vmbus_channel *c)
c->low_latency = false;
}
+bool dev_is_vmbus(struct device *dev);
+
void vmbus_onmessage(void *context);
int vmbus_request_offers(void);
--
1.7.4.1
^ permalink raw reply related [flat|nested] 23+ messages in thread* Re: [PATCH V2 15/15] hyperv: Add a function to detect if the device is a vmbus dev
2016-12-03 20:34 ` [PATCH V2 15/15] hyperv: Add a function to detect if the device is a vmbus dev kys
@ 2016-12-06 10:54 ` Greg KH
2016-12-06 14:58 ` KY Srinivasan
0 siblings, 1 reply; 23+ messages in thread
From: Greg KH @ 2016-12-06 10:54 UTC (permalink / raw)
To: kys
Cc: linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara, Haiyang Zhang
On Sat, Dec 03, 2016 at 12:34:42PM -0800, kys@exchange.microsoft.com wrote:
> From: Haiyang Zhang <haiyangz@microsoft.com>
>
> On Hyper-V, every VF interface has a corresponding synthetic
> interface managed by netvsc that share the same MAC
> address. netvsc registers for netdev events to manage this
> association. Currently we use the MAC address to manage this
> association but going forward, we want to use a serial number that
> the host publishes. To do this we need functionality similar
> to dev_is_pci() for the vmbus devices. Implement such a function.
>
> This function will be used in subsequent patches to netvsc and in the
> interest of eliminating cross tree dependencies, this patch is being
> submitted first.
As I stated before, I want to see those patches before I will take this
one...
thanks,
greg k-h
^ permalink raw reply [flat|nested] 23+ messages in thread
* RE: [PATCH V2 15/15] hyperv: Add a function to detect if the device is a vmbus dev
2016-12-06 10:54 ` Greg KH
@ 2016-12-06 14:58 ` KY Srinivasan
0 siblings, 0 replies; 23+ messages in thread
From: KY Srinivasan @ 2016-12-06 14:58 UTC (permalink / raw)
To: Greg KH
Cc: linux-kernel, devel, olaf, apw, vkuznets, jasowang,
leann.ogasawara, Haiyang Zhang
> -----Original Message-----
> From: Greg KH [mailto:gregkh@linuxfoundation.org]
> Sent: Tuesday, December 6, 2016 2:54 AM
> To: KY Srinivasan <kys@microsoft.com>
> Cc: linux-kernel@vger.kernel.org; devel@linuxdriverproject.org;
> olaf@aepfle.de; apw@canonical.com; vkuznets@redhat.com;
> jasowang@redhat.com; leann.ogasawara@canonical.com; Haiyang Zhang
> <haiyangz@microsoft.com>
> Subject: Re: [PATCH V2 15/15] hyperv: Add a function to detect if the device
> is a vmbus dev
>
> On Sat, Dec 03, 2016 at 12:34:42PM -0800, kys@exchange.microsoft.com
> wrote:
> > From: Haiyang Zhang <haiyangz@microsoft.com>
> >
> > On Hyper-V, every VF interface has a corresponding synthetic
> > interface managed by netvsc that share the same MAC
> > address. netvsc registers for netdev events to manage this
> > association. Currently we use the MAC address to manage this
> > association but going forward, we want to use a serial number that
> > the host publishes. To do this we need functionality similar
> > to dev_is_pci() for the vmbus devices. Implement such a function.
> >
> > This function will be used in subsequent patches to netvsc and in the
> > interest of eliminating cross tree dependencies, this patch is being
> > submitted first.
>
> As I stated before, I want to see those patches before I will take this
> one...
Thanks Greg; we will send you the patches.
K. Y
>
> thanks,
>
> greg k-h
^ permalink raw reply [flat|nested] 23+ messages in thread
* Re: [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg()
2016-12-03 20:34 ` [PATCH V2 01/15] Drivers: hv: vmbus: Raise retry/wait limits in vmbus_post_msg() kys
` (13 preceding siblings ...)
2016-12-03 20:34 ` [PATCH V2 15/15] hyperv: Add a function to detect if the device is a vmbus dev kys
@ 2016-12-06 10:07 ` Greg KH
14 siblings, 0 replies; 23+ messages in thread
From: Greg KH @ 2016-12-06 10:07 UTC (permalink / raw)
To: kys; +Cc: linux-kernel, devel, olaf, apw, vkuznets, jasowang, leann.ogasawara
On Sat, Dec 03, 2016 at 12:34:28PM -0800, kys@exchange.microsoft.com wrote:
> From: Vitaly Kuznetsov <vkuznets@redhat.com>
>
> DoS protection conditions were altered in WS2016 and now it's easy to get
> -EAGAIN returned from vmbus_post_msg() (e.g. when we try changing MTU on a
> netvsc device in a loop). All vmbus_post_msg() callers don't retry the
> operation and we usually end up with a non-functional device or crash.
>
> While host's DoS protection conditions are unknown to me my tests show that
> it can take up to 10 seconds before the message is sent so doing udelay()
> is not an option, we really need to sleep. Almost all vmbus_post_msg()
> callers are ready to sleep but there is one special case:
> vmbus_initiate_unload() which can be called from interrupt/NMI context and
> we can't sleep there. I'm also not sure about the lonely
> vmbus_send_tl_connect_request() which has no in-tree users but its external
> users are most likely waiting for the host to reply so sleeping there is
> also appropriate.
>
> Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
> Signed-off-by: K. Y. Srinivasan <kys@microsoft.com>
Shouldn't this go to stable kernels so that 4.9 and earlier can work
properly on WS2016?
thanks,
greg k-h
^ permalink raw reply [flat|nested] 23+ messages in thread