, October 11, 2014 3:36 AM
To: Long Li
Cc: David Miller; o...@aepfle.de; net...@vger.kernel.org; jasow...@redhat.com;
linux-kernel@vger.kernel.org; a...@canonical.com; de...@linuxdriverproject.org
Subject: Re: [PATCH net 1/1] hyperv: Fix a bug in netvsc_send()
On Fri, Oct 10, 2014 at 11:39:00PM +
Srinivasan; Haiyang Zhang; Long Li; net...@vger.kernel.org;
linux-kernel@vger.kernel.org
Subject: skbuff_fclone_cache poison overwritten
While using 3.18.0-rc3.x86_64-00116-g6ac94d3 on a Hyper-V 2012 R2 the poison in
skbuff_fclone_cache was overwritten:
[39099.484435] sd 7:0:0:0: [sdi] Attached SCSI
Sitsofe, can you try the patch attached to see if it helps with the problem?
Long
-Original Message-
From: linux-kernel-ow...@vger.kernel.org
[mailto:linux-kernel-ow...@vger.kernel.org] On Behalf Of Peter Zijlstra
Sent: Monday, November 10, 2014 1:44 AM
To: Sitsofe Wheeler
Cc: KY
added.
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Reviewed-by: Long Li lon...@microsoft.com
---
drivers/scsi/storvsc_drv.c | 26 ++
1 files changed, 26 insertions(+), 0 deletions(-)
diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c index
james.bottom...@hansenpartnership.com
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Reviewed-by: Long Li lon...@microsoft.com
---
drivers/scsi/storvsc_drv.c | 13 +
1 files changed, 13 insertions(+), 0 deletions(-)
diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi
.
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Reviewed-by: Long Li lon...@microsoft.com
---
drivers/scsi/storvsc_drv.c | 19 +++
1 files changed, 7 insertions(+), 12 deletions(-)
diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c index
e3ba251
these new limits. Fix this bug by setting the limits based on the host.
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Reviewed-by: Long Li lon...@microsoft.com
---
drivers/scsi/storvsc_drv.c | 15 ---
1 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/drivers/scsi
Thanks Martin for the explanation.
I'll send out another patch.
-Original Message-
From: Martin K. Petersen [mailto:martin.peter...@oracle.com]
Sent: Thursday, December 11, 2014 7:04 PM
To: Long Li
Cc: Martin K. Petersen; KY Srinivasan; Haiyang Zhang;
jbottom...@parallels.com
:29PM -0400, David Miller wrote:
From: K. Y. Srinivasan k...@microsoft.com
Date: Sun, 5 Oct 2014 10:42:51 -0700
After the packet is successfully sent, we should not touch the
packet as it may have been freed. This patch is based on the work
done by Long Li lon...@microsoft.com
When ring buffer returns an error indicating retry, storvsc may not return a
proper error code to SCSI when bounce buffer is not used. This has introduced
I/O freeze on RAID running atop storvsc devices. This patch fixes it by always
returning a proper error code.
Signed-off-by: Long Li lon
Signed-off-by: Long Li lon...@microsoft.com
---
drivers/scsi/scsi_devinfo.c | 1 +
drivers/scsi/storvsc_drv.c | 10 --
2 files changed, 1 insertion(+), 10 deletions(-)
diff --git a/drivers/scsi/scsi_devinfo.c b/drivers/scsi/scsi_devinfo.c
index c1d04d4..f6fe0b2 100644
--- a/drivers/scsi
Long == Long Li lon...@microsoft.com writes:
Long MSFT targets currently claim SPC-2 compliance while they implement
Long post SPC-2 features. With this patch we can correctly handle
Long WRITE_SAME_16 issues.
Handle the issues or handle WRITE SAME(10/16)?
With this patch, the SCSI
-Original Message-
From: Sitsofe Wheeler [mailto:sits...@gmail.com]
Sent: Wednesday, December 10, 2014 12:58 PM
To: Long Li
Cc: KY Srinivasan; Haiyang Zhang; jbottom...@parallels.com; linux-
s...@vger.kernel.org; de...@linuxdriverproject.org; linux-
ker...@vger.kernel.org
Subject
to not make this assumption.
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Reviewed-by: Long Li lon...@microsoft.com
---
drivers/scsi/storvsc_drv.c | 98 +--
1 files changed, 57 insertions(+), 41 deletions(-)
diff --git a/drivers/scsi
have acquired. Fix the bug.
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Reviewed-by: Long Li lon...@microsoft.com
Cc: sta...@vger.kernel.org
---
drivers/scsi/storvsc_drv.c | 15 ---
1 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/drivers/scsi
keith.ma...@microsoft.com
Rather than look for sets of specific protocol versions, make decisions based
on
ranges. This will be safer and require fewer changes going forward as we add
more storage protocol versions.
Reviewed-by: Long Li lon...@microsoft.com
Tested-by: Alex Ng ale
value to track protocol versions to simplify comparisons and to
be
consistent with vmbus version tracking.
Reviewed-by: Long Li lon...@microsoft.com
Tested-by: Alex Ng ale...@microsoft.com
Signed-off-by: Keith Mange keith.ma...@microsoft.com
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Use storage protocol version instead of vmbus protocol version when
determining storage capabilities.
Reviewed-by: Long Li lon...@microsoft.com
Tested-by: Alex Ng ale...@microsoft.com
Signed-off-by: Keith Mange keith.ma...@microsoft.com
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
Use correct defaults for values determined by protocol negotiation, instead of
resetting them with every scsi controller.
Reviewed-by: Long Li lon...@microsoft.com
Tested-by: Alex Ng ale...@microsoft.com
Signed-off-by: Keith Mange keith.ma...@microsoft.com
Signed-off-by: K. Y. Srinivasan k
...@microsoft.com
Currently we are making decisions based on vmbus protocol versions that have
been negotiated; use storage potocol versions instead.
Reviewed-by: Long Li lon...@microsoft.com
Tested-by: Alex Ng ale...@microsoft.com
Signed-off-by: Keith Mange keith.ma...@microsoft.com
Signed-off-by: K. Y
for Windows10 and above hosts.
Reviewed-by: Long Li lon...@microsoft.com
Tested-by: Alex Ng ale...@microsoft.com
Signed-off-by: Keith Mange keith.ma...@microsoft.com
Signed-off-by: K. Y. Srinivasan k...@microsoft.com
---
drivers/scsi/storvsc_drv.c |6 +-
1 files changed, 5 insertions(+), 1
-Original Message-
From: KY Srinivasan
Sent: Friday, July 03, 2015 11:35 AM
To: Vitaly Kuznetsov; linux-s...@vger.kernel.org
Cc: Long Li; Haiyang Zhang; James E.J. Bottomley;
de...@linuxdriverproject.org;
linux-kernel@vger.kernel.org
Subject: RE: [PATCH] scsi: storvsc: use
Introduce a logging level for storvsc to log certain error/warning messages.
Those messages are helpful in some environments, e.g. Microsoft Azure, for
customer support and troubleshooting purposes.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/scsi/storvsc_drv.
> -Original Message-
> From: Vitaly Kuznetsov [mailto:vkuzn...@redhat.com]
> Sent: Friday, December 4, 2015 1:53 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; James E.J. B
Introduce a logging level for storvsc to log certain error/warning messages.
Those messages are helpful in some environments, e.g. Microsoft Azure, for
customer support and troubleshooting purposes.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/scsi/storvsc_drv.
Thanks Joe.
I'll send out another patch.
> -Original Message-
> From: Joe Perches [mailto:j...@perches.com]
> Sent: Thursday, December 3, 2015 6:28 PM
> To: Long Li <lon...@microsoft.com>; KY Srinivasan <k...@microsoft.com>;
> Haiyang Zhang <haiya...@micros
une 6, 2016 8:42 PM
> To: Long Li <lon...@microsoft.com>
> Cc: Tom Yan <tom.t...@gmail.com>; James E.J. Bottomley
> <j...@linux.vnet.ibm.com>; Martin K. Petersen
> <martin.peter...@oracle.com>; linux-s...@vger.kernel.org; linux-
> ker...@vger.ke
Hi Martin,
Yes I would like to test it. Please let me know when they get in there.
Thanks
Long
> -Original Message-
> From: Martin K. Petersen [mailto:martin.peter...@oracle.com]
> Sent: Tuesday, May 31, 2016 7:44 PM
> To: Long Li <lon...@microsoft.com>
>
ing values on optional VPD page 0xb0 "Block
Limits".
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/scsi/sd.c | 7 ++-
1 file changed, 2 insertions(+), 5 deletions(-)
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index 60bff78..d8c4047 100644
--- a/driv
e it again. If
they want max_sectors to be set by sd, they can use BLOCK LIMITS VPD to tell it
to do so.
> -Original Message-
> From: Tom Yan [mailto:tom.t...@gmail.com]
> Sent: Saturday, June 4, 2016 1:41 AM
> To: Long Li <lon...@microsoft.com>
> Cc: James E.J. Bot
The block sector size should be in unit of 512 bytes, not in bytes.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/scsi/sd.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index 428c03e..4bce17e 100644
--- a/d
> -Original Message-
> From: Bart Van Assche [mailto:bart.vanass...@sandisk.com]
> Sent: Thursday, May 26, 2016 7:19 PM
> To: Long Li <lon...@microsoft.com>; James E.J. Bottomley
> <j...@linux.vnet.ibm.com>; Martin K. Petersen
> <martin.peter...@o
> -Original Message-
> From: Dexuan Cui
> Sent: Sunday, January 15, 2017 7:12 PM
> To: Long Li <lon...@microsoft.com>; KY Srinivasan <k...@microsoft.com>;
> Haiyang Zhang <haiya...@microsoft.com>
> Cc: de...@linuxdriverproject.org; linux-kernel@vger.
status
hv_pcibus_removed, we can avoid this situation.
Signed-off-by: Long Li <lon...@microsoft.com>
Reported-by: Xiaofeng Wang <xiaof...@redhat.com>
Acked-by: K. Y. Srinivasan <k...@microsoft.com>
---
drivers/pci/host/pci-hyperv.c | 20 +---
1 file changed, 1
<de...@microsoft.com> for pointing out the race condition in
create_root_hv_pci_bus().
Signed-off-by: Long Li <lon...@microsoft.com>
Reported-by: Xiaofeng Wang <xiaof...@redhat.com>
Acked-by: K. Y. Srinivasan <k...@microsoft.com>
---
drivers/pci/host/pci-hyperv.c | 4
> Sent: Friday, January 27, 2017 10:42 AM
> To: Long Li <lon...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn Helgaas <bhelg...@google.com>
> Cc: de...@linuxdriverproject.org; linux-...@vger.kernel.org; linux-
> ker...@vger.kernel.org; Long Li <lo
> -Original Message-
> From: Bjorn Helgaas [mailto:bhelg...@google.com]
> Sent: Saturday, February 11, 2017 9:35 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; de..
status hv_pcibus_removed, we can avoid this situation.
Signed-off-by: Long Li <lon...@microsoft.com>
Reported-by: Xiaofeng Wang <xiaof...@redhat.com>
Acked-by: K. Y. Srinivasan <k...@microsoft.com>
---
drivers/pci/host/pci-hyperv.c | 20 +---
1 file changed, 1
<de...@microsoft.com> for pointing out the race condition in
create_root_hv_pci_bus().
Signed-off-by: Long Li <lon...@microsoft.com>
Reported-by: Xiaofeng Wang <xiaof...@redhat.com>
Acked-by: K. Y. Srinivasan <k...@microsoft.com>
---
drivers/pci/host/pci-hyperv.c | 4
Ok, I will resend.
> -Original Message-
> From: Greg KH [mailto:g...@kroah.com]
> Sent: Saturday, February 25, 2017 12:02 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn
From: Long Li <lon...@microsoft.com>
When sense message is present on error, we should pass along to the upper layer
to decide how to deal with the error. This patch fixes connectivity issues with
Fiber Channel devices.
---
drivers/scsi/storvsc_drv.c | 3 +++
1 file changed, 3 inse
From: Long Li <lon...@microsoft.com>
This patch set fixes connectivity issues and improves performance for Fiber
Channel disks.
Long Li (3):
Use tagged SRB requests if supported by the device
Properly handle SRB_ERROR when sense message is present
Use block layer default segmen
From: Long Li <lon...@microsoft.com>
We no long have the restriction of page size limit in the SG list. Remove it.
The driver can properly handle default block segment size.
---
drivers/scsi/storvsc_drv.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/scsi/storvsc_d
From: Long Li <lon...@microsoft.com>
Properly set SRB flags when hosting device supports tagged queuing. This patch
improves the performance on Fiber Channel disks.
---
drivers/scsi/storvsc_drv.c | 8
1 file changed, 8 insertions(+)
diff --git a/drivers/scsi/storvsc_drv.c b/d
> -Original Message-
> From: Dexuan Cui
> Sent: Tuesday, September 13, 2016 2:51 AM
> To: Long Li <lon...@microsoft.com>; KY Srinivasan <k...@microsoft.com>;
> Haiyang Zhang <haiya...@microsoft.com>; Bjorn Helgaas
> <bhelg...@google.com>
> C
> -Original Message-
> From: devel [mailto:driverdev-devel-boun...@linuxdriverproject.org] On
> Behalf Of Long Li
> Sent: Tuesday, September 13, 2016 10:33 AM
> To: Dexuan Cui <de...@microsoft.com>; KY Srinivasan
> <k...@microsoft.com>; Haiyang Zhang &
> -Original Message-
> From: Dexuan Cui
> Sent: Tuesday, September 13, 2016 10:45 PM
> To: Long Li <lon...@microsoft.com>; KY Srinivasan <k...@microsoft.com>;
> Haiyang Zhang <haiya...@microsoft.com>; Bjorn Helgaas
> <bhelg...@google.com>
> C
From: Long Li <lon...@microsoft.com>
A PCI_EJECT message can arrive at the same time we are calling
pci_scan_child_bus in the workqueue for the previous PCI_BUS_RELATIONS message
or in create_root_hv_pci_bus(), in this case we could potentailly modify the
bus from multiple places. Pr
From: Long Li <lon...@microsoft.com>
hv_pci_devices_present is called in hv_pci_remove when we remove a PCI device
from host (e.g. by disabling SRIOV on a device). In hv_pci_remove, the bus is
already removed before the call, so we don't need to rescan the bus in the
workqueue schedule
From: Long Li <lon...@microsoft.com>
hv_pci_devices_present is called in hv_pci_remove when we remove a PCI device
from host (e.g. by disabling SRIOV on a device). In hv_pci_remove, the bus is
already removed before the call, so we don't need to rescan the bus in the
workqueue schedule
From: Long Li <lon...@microsoft.com>
A PCI_EJECT message can arrive at the same time we are calling
pci_scan_child_bus in the workqueue for the previous PCI_BUS_RELATIONS message,
in this case we could potentailly modify the bus from two places. Properly lock
the bus access.
Sign
From: Long Li <lon...@microsoft.com>
The host keeps sending heartbeat packets independent of guest responding to
them. In some situations, there might be multiple heartbeat packets pending in
the ring buffer. Don't lose them, read them all.
Signed-off-by: Long Li <lon...@micr
> -Original Message-
> From: Johannes Thumshirn [mailto:jthumsh...@suse.de]
> Sent: Wednesday, September 7, 2016 12:47 AM
> To: Long Li <lon...@exchange.microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsof
> -Original Message-
> From: Bjorn Helgaas [mailto:helg...@kernel.org]
> Sent: Tuesday, September 27, 2016 12:30 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn
> -Original Message-
> From: Vitaly Kuznetsov [mailto:vkuzn...@redhat.com]
> Sent: Thursday, September 29, 2016 2:22 AM
> To: KY Srinivasan <k...@microsoft.com>; Long Li <lon...@microsoft.com>
> Cc: Haiyang Zhang <haiya...@microsoft.com>;
> de...@
From: Long Li <lon...@microsoft.com>
hv_pci_devices_present is called in hv_pci_remove when we remove a PCI device
from host (e.g. by disabling SRIOV on a device). In hv_pci_remove, the bus is
already removed before the call, so we don't need to rescan the bus in the
workqueue schedule
From: Long Li <lon...@microsoft.com>
A PCI_EJECT message can arrive at the same time we are calling
pci_scan_child_bus in the workqueue for the previous PCI_BUS_RELATIONS message
or in create_root_hv_pci_bus(), in this case we could potentailly modify the
bus from multiple places. Pr
From: Long Li <lon...@microsoft.com>
hv_do_hypercall assumes that we pass a segment from a physically
continuous buffer. Buffer allocated on the stack may not work if
CONFIG_VMAP_STACK=y is set.
Change to use kmalloc to allocate this buffer.
The v2 patch adds locking to access the pre-all
> -Original Message-
> From: Greg KH [mailto:gre...@linuxfoundation.org]
> Sent: Monday, November 7, 2016 11:00 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn
> -Original Message-
> From: devel [mailto:driverdev-devel-boun...@linuxdriverproject.org] On
> Behalf Of Long Li
> Sent: Tuesday, November 8, 2016 8:57 AM
> To: Greg KH <gre...@linuxfoundation.org>
> Cc: linux-...@vger.kernel.org; Haiyang Zhang <haiya...@mic
From: Long Li <lon...@microsoft.com>
hv_do_hypercall assumes that we pass a segment from a physically continuous
buffer. Buffer allocated on the stack may not work if CONFIG_VMAP_STACK=y is
set. Use kmalloc to allocate this buffer.
Signed-off-by: Long Li <lon...@microsoft.com&
From: Long Li <lon...@microsoft.com>
We need to pass a segment from a physically continuous buffer to
hv_do_hypercall. Buffer allocated on the stack may not work if
CONFIG_VMAP_STACK=y is set. Moving the params buffer from stack to buffer
returned by kmalloc.
Signed-off-by: Long L
Hi Bjorn
This patch is still pending. The patch has been ack'ed.
Do you want me to resend this patch?
Thanks
Long
> -Original Message-
> From: KY Srinivasan
> Sent: Friday, November 11, 2016 2:21 PM
> To: Bjorn Helgaas <helg...@kernel.org>; Long Li <lon...@microso
Hi Bjorn,
The patch is still pending (along with 1/2 v3). Please let me know if you want
me to resend the two patches.
Thanks
Long
> -Original Message-
> From: KY Srinivasan
> Sent: Tuesday, October 4, 2016 1:49 PM
> To: Long Li <lon...@microsoft.com>; Haiy
> -Original Message-
> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> Sent: Monday, December 5, 2016 8:53 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn
From: Long Li <lon...@microsoft.com>
The ring buffer code uses %= to calculate index. For x86/64, %= compiles to
div, more than 10 times slower than sub.
Replace div with sub for this data heavy code path.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/hv/ring_
From: Long Li <lon...@microsoft.com>
Hyper-v host guarantees that a hypercall will finish in reasonable time.
Retry infinitely on transient failures to avoid returning error to upper layer.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/hv/conne
From: Long Li <lon...@microsoft.com>
Hyper-v host guarantees that a hypercall will succeed. Retry infinitely to
avoid returning transient failures to upper layer.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/hv/connection.c | 17 -
1 file changed,
> -Original Message-
> From: Greg KH [mailto:g...@kroah.com]
> Sent: Wednesday, January 4, 2017 12:51 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; de...@linuxdriverproject.
> -Original Message-
> From: Dan Carpenter [mailto:dan.carpen...@oracle.com]
> Sent: Wednesday, January 4, 2017 1:48 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; de...@
> -Original Message-
> From: Greg KH [mailto:g...@kroah.com]
> Sent: Wednesday, January 04, 2017 11:48 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; de...@linuxdriverproject.
> -Original Message-
> From: Greg KH [mailto:g...@kroah.com]
> Sent: Friday, January 06, 2017 11:43 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; de...@linuxdriverproject.
> -Original Message-
> From: Dan Carpenter [mailto:dan.carpen...@oracle.com]
> Sent: Thursday, January 05, 2017 3:40 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; de...@
From: Long Li <lon...@microsoft.com>
hv_pci_devices_present is called in hv_pci_remove when we remove a PCI
device from host (e.g. by disabling SRIOV on a device). In hv_pci_remove,
the bus is already removed before the call, so we don't need to rescan the
bus in the workqueue schedule
From: Long Li <lon...@microsoft.com>
A PCI_EJECT message can arrive at the same time we are calling
pci_scan_child_bus in the workqueue for the previous PCI_BUS_RELATIONS
message or in create_root_hv_pci_bus(), in this case we could potentially
modify the bus from multiple places.
Properl
From: Long Li <lon...@microsoft.com>
The host may send multiple negotiation packets (due to timeout) before the
KVP user-mode daemon is connected. We need to defer processing those packets
until the daemon is negotiated and connected. It's okay for guest to respond
to all negotiation p
> -Original Message-
> From: Vitaly Kuznetsov [mailto:vkuzn...@redhat.com]
> Sent: Thursday, March 23, 2017 9:04 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Stephen Hemm
From: Long Li <lon...@microsoft.com>
The host may send multiple negotiation packets (due to timeout) before the
KVP user-mode daemon is connected. We need to defer processing those packets
until the daemon is negotiated and connected. It's okay for guest to respond
to all negotiation p
omic bug
>
> From: K. Y. Srinivasan <k...@microsoft.com>
>
> The memory allocation here needs to be non-blocking.
> Fix the issue.
>
> Signed-off-by: K. Y. Srinivasan <k...@microsoft.com>
> Cc: <sta...@vger.kernel.org>
Reviewed-by: Long Li <lon...@mi
in specifying CPU affinity
>
> From: K. Y. Srinivasan <k...@microsoft.com>
>
> When we have 32 or more CPUs in the affinity mask, we should use a special
> constant to specify that to the host. Fix this issue.
>
> Signed-off-by: K. Y. Srinivasan <k...@mic
<vkuzn...@redhat.com>.
Signed-off-by: Long Li <lon...@microsoft.com>
---
drivers/hv/hv_kvp.c | 12 +++-
1 file changed, 7 insertions(+), 5 deletions(-)
diff --git a/drivers/hv/hv_kvp.c b/drivers/hv/hv_kvp.c
index de26371..845b70b 100644
--- a/drivers/hv/hv_kvp.c
+++ b/d
> -Original Message-
> From: Bjorn Helgaas [mailto:helg...@kernel.org]
> Sent: Thursday, March 16, 2017 1:07 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn
> -Original Message-
> From: Long Li
> Sent: Sunday, March 19, 2017 7:49 PM
> To: 'Vitaly Kuznetsov' <vkuzn...@redhat.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Stephen Hemminger
> <sthem...@microsoft
> -Original Message-
> From: Vitaly Kuznetsov [mailto:vkuzn...@redhat.com]
> Sent: Friday, March 17, 2017 9:16 AM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Stephen Hemm
The host may send multiple KVP packets before the negotiation with daemon
is finished. We need to keep those packets in ring buffer until the daemon
is negotiated and connected.
This patch is based on the work of Nick Meier <nick.me...@microsoft.com>
Signed-off-by: Long Li <lon...@micr
> -Original Message-
> From: Bjorn Helgaas [mailto:helg...@kernel.org]
> Sent: Thursday, March 16, 2017 1:07 PM
> To: Long Li <lon...@microsoft.com>
> Cc: KY Srinivasan <k...@microsoft.com>; Haiyang Zhang
> <haiya...@microsoft.com>; Bjorn
> -Original Message-
> From: Bart Van Assche [mailto:bart.vanass...@sandisk.com]
> Sent: Wednesday, April 5, 2017 5:32 PM
> To: linux-kernel@vger.kernel.org; linux-bl...@vger.kernel.org; Long Li
> <lon...@microsoft.com>; ax...@kernel.dk
> Cc: Stephen Hemminger
> -Original Message-
> From: KY Srinivasan
> Sent: Wednesday, April 5, 2017 9:21 PM
> To: Bart Van Assche <bart.vanass...@sandisk.com>; linux-
> ker...@vger.kernel.org; linux-bl...@vger.kernel.org; Long Li
> <lon...@microsoft.com>; ax...@kernel.dk
> -Original Message-
> From: Bart Van Assche [mailto:bart.vanass...@sandisk.com]
> Sent: Monday, April 10, 2017 4:48 PM
> To: linux-kernel@vger.kernel.org; linux-bl...@vger.kernel.org; KY Srinivasan
> <k...@microsoft.com>; Long Li <lon...@microsoft.com>; ax
From: Long Li <lon...@microsoft.com>
Under heavy I/O, one hardware queue may be unable to dispatch any I/O to the
device layer. This poses a problem with restarting this hardware queue on I/O
finish in blk_mq_sched_restart_queues(), becaue there is nothing pending that
will finish in
From: Long Li <lon...@microsoft.com>
Similar to sending transfer message with page payload, this function creates a
SMBD data packet and send it over to RDMA, from iov passed from upper layer.
Signed-off-by: Long Li <lon...@microsoft.com>
---
fs/cifs/cifs
From: Long Li <lon...@microsoft.com>
Implement the code for connecting to SMBD server. The client and server are
connected using RC Queue Pair over RDMA API, which suppports Infiniband, RoCE
and iWARP. Upper layer code can call cifs_create_rdma_session to establish a
SMBD RDMA conn
From: Long Li <lon...@microsoft.com>
Define a new structure for SMBD transport. This stucture will have all the
information on the transport, and it will be stored in the current SMB session.
Signed-off-by: Long Li <lon...@microsoft.com>
---
fs/cifs/cif
From: Long Li <lon...@microsoft.com>
Define the packet format for a SMBD data packet with payload
Signed-off-by: Long Li <lon...@microsoft.com>
---
fs/cifs/cifsrdma.h | 15 +++
1 file changed, 15 insertions(+)
diff --git a/fs/cifs/cifsrdma.h b/fs/cifs/cifsrdma.h
i
From: Long Li <lon...@microsoft.com>
For each channel, SMBD defines per-channel parameters. Those can be negotiated
with the server, and are restricted by RDMA hardware limitations.
Signed-off-by: Long Li <lon...@microsoft.com>
---
fs/cifs/cifsrdma.c | 14 ++
fs/cif
From: Long Li <lon...@microsoft.com>
SMB client and server maintain a credit system on the SMBD transport. Credits
are used to tell when the client or server can send a packet to the peer, based
on current memory or resource usage.
Signed-off-by: Long Li <lon...@microsoft.com>
From: Long Li <lon...@microsoft.com>
SMB3 defines a protocol for transfer data over RDMA transport such as
Infiniband, RoCE and iWARP. The prococol is published in [MS-SMBD]
(https://msdn.microsoft.com/en-us/library/hh536346.aspx).
This is the series 1 of two patch sets. This pat
From: Long Li <lon...@microsoft.com>
Define the data structure for receiving a SMBD response. SMBD responses are
allocated through a per-channel mempool. For each server initiated response
message, the SMB client must post a SMBD response to the local hardware.
Signed-off-by: Long L
From: Long Li <lon...@microsoft.com>
When doing mount with "-o rdma", user can specify this is for connecting to a
SMBD session.
Signed-off-by: Long Li <lon...@microsoft.com>
---
fs/cifs/cifs_debug.c | 2 ++
fs/cifs/cifsfs.c | 2 ++
fs/cifs/cifsglob.h | 2 ++
fs/
From: Long Li <lon...@microsoft.com>
Implement cifs_rdma_write for send an upper layer data. Upper layer uses this
function to do a RDMA send. This function is also used to pass SMB packets for
doing a RDMA read/write via memory registration.
Signed-off-by: Long Li <lon...@micr
1 - 100 of 1071 matches
Mail list logo