On Wed, Nov 16, 2016 at 06:00:00PM +, Harry van Haaren wrote:
> This series of RFC patches implements the libeventdev API and a software
> eventdev PMD.
>
> The implementation here is intended to enable the community to use the
> eventdev API, specifically to test if the API serves the
It is good that RC1 on Jan.11th is a hard deadline. We also need ensure that
RC1 is a complete package with all features. If RC1 is out but missing some big
features, then the test results on RC1 vs RC2 may have big difference.
So, could we ensure that RC1 is a complete feature package, if any
DPDK has been inherently a PCI inclined framework. Because of this, the
design of device tree (or list) within DPDK is also PCI inclined. A non-PCI
device doesn't have a way of being expressed without using hooks started from
EAL to PMD.
With this cover letter, some patches are presented which
From: Jan Viktorin
Signed-off-by: Jan Viktorin
Signed-off-by: Shreyansh Jain
---
lib/librte_eal/common/include/rte_common.h | 18 ++
1 file changed, 18 insertions(+)
diff --git a/lib/librte_eal/common/include/rte_common.h
A device is connected to a bus and services by a driver associated with
the bus. It is responsibility of the bus to identify the devices (scan)
and then assign each device to a matching driver.
A PMD would allocate a rte_xxx_driver and rte_xxx_device.
rte_xxx_driver has rte_driver and rte_bus
A bus is managed using a 'bus driver'. This patch introduces a sample
PCI bus driver which essentially is the scan and match implementation for
PCI.
There are multiple possible alternatives to where such a driver can be
kept within the DPDK code:
1. librte_bus
- For each type of bus, there would
Primary changes done by this patch are based on:
- Devices belong to the bus hence the device list is bus instance
specific
- Similarly, drivers belong to the bus and thus they too are enclosed
within the bus instance.
- All device insertion and driver registration should proceed through
This patch makes necessary changes to the EAL for handling bus scan and
probe. Most of the scan function has been moved to bus/* implementation,
(currently only for pci, linuxapp). There are still some functions which
exists in the EAL (bind, unbind module, mmap etc).
Missing/Grey area;
- Should
This patch demonstrates how eth_driver can be replaced with appropriate
changes for rte_xxx_driver from the PMD itself. It uses ixgbe_ethernet as
an example.
A large set of changes exists in the rte_ethdev.c - primarily because too
much PCI centric code (names, assumption of rte_pci_device) still
In function pci_mknod_uio_dev() in lib/librte_eal/eal/eal_pci_uio.c,
The return value of mknod() is ret, not f got by fopen().
So the value of ret should be checked for mknod().
Fixes: f7f97c16048e ("pci: add option --create-uio-dev to run without hotplug")
Signed-off-by: Wei Dai
---
v3:
*
As usaual, sorry for late response :/
On Thu, Oct 13, 2016 at 08:50:52PM +0300, Michael S. Tsirkin wrote:
> Hi!
> So it looks like we face a problem with cross-version
> migration when using vhost. It's not new but became more
> acute with the advent of vhost user.
>
> For users to be able to
On 11/17/2016 09:29 AM, Yuanhan Liu wrote:
> As usaual, sorry for late response :/
>
> On Thu, Oct 13, 2016 at 08:50:52PM +0300, Michael S. Tsirkin wrote:
>> Hi!
>> So it looks like we face a problem with cross-version
>> migration when using vhost. It's not new but became more
>> acute with the
Repost from the moving at dpdk.org mailing list to get a wider audience.
Original thread: http://dpdk.org/ml/archives/moving/2016-November/59.html
Hi,
I'd like to propose a change to the DPDK committer model. Currently we have one
committer for the master branch of the DPDK project.
One
Signed-off-by: Qiming Yang
---
drivers/net/e1000/igb_ethdev.c | 46 ++
1 file changed, 46 insertions(+)
diff --git a/drivers/net/e1000/igb_ethdev.c b/drivers/net/e1000/igb_ethdev.c
index 2fddf0c..c47f1c8 100644
--- a/drivers/net/e1000/igb_ethdev.c
+++
This patch added API for 'rte_eth_dev_fwver_get'
void rte_eth_dev_fwver_get(uint8_t port_id,
char *fw_version, int fw_length);
Signed-off-by: Qiming Yang
---
lib/librte_ether/rte_ethdev.c | 12
lib/librte_ether/rte_ethdev.h | 18 ++
This patch enhances the ethtool-API to support to show the
bus information and firmware version as same as kernel
version ethtool displayed.
Signed-off-by: Qiming Yang
---
examples/ethtool/ethtool-app/ethapp.c | 2 ++
examples/ethtool/lib/rte_ethtool.c| 3 +++
2 files changed, 5
Signed-off-by: Qiming Yang
---
drivers/net/i40e/i40e_ethdev.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
index 67778ba..f79bc5e 100644
--- a/drivers/net/i40e/i40e_ethdev.c
+++
Signed-off-by: Qiming Yang
---
drivers/net/ixgbe/ixgbe_ethdev.c | 18 ++
1 file changed, 18 insertions(+)
diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c
index edc9b22..da9aa31 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.c
+++
Signed-off-by: Qiming Yang
---
drivers/net/e1000/igb_ethdev.c | 46 ++
1 file changed, 46 insertions(+)
diff --git a/drivers/net/e1000/igb_ethdev.c b/drivers/net/e1000/igb_ethdev.c
index 2fddf0c..c47f1c8 100644
--- a/drivers/net/e1000/igb_ethdev.c
+++
Signed-off-by: Qiming Yang
---
drivers/net/ixgbe/ixgbe_ethdev.c | 18 ++
1 file changed, 18 insertions(+)
diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c b/drivers/net/ixgbe/ixgbe_ethdev.c
index edc9b22..da9aa31 100644
--- a/drivers/net/ixgbe/ixgbe_ethdev.c
+++
Signed-off-by: Qiming Yang
---
drivers/net/i40e/i40e_ethdev.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/drivers/net/i40e/i40e_ethdev.c b/drivers/net/i40e/i40e_ethdev.c
index 67778ba..f79bc5e 100644
--- a/drivers/net/i40e/i40e_ethdev.c
+++
This patch enhances the ethtool-API to support to show the
bus information and firmware version as same as kernel
version ethtool displayed.
Signed-off-by: Qiming Yang
---
examples/ethtool/ethtool-app/ethapp.c | 2 ++
examples/ethtool/lib/rte_ethtool.c| 3 +++
2 files changed, 5
On Thu, Nov 17, 2016 at 09:47:09AM +0100, Maxime Coquelin wrote:
>
>
> On 11/17/2016 09:29 AM, Yuanhan Liu wrote:
> >As usaual, sorry for late response :/
> >
> >On Thu, Oct 13, 2016 at 08:50:52PM +0300, Michael S. Tsirkin wrote:
> >>Hi!
> >>So it looks like we face a problem with cross-version
The list of segments to free was wrongly manipulated ending by only freeing
the first segment instead of freeing all of them. The last one still
belongs to the NIC and thus should not be freed.
Fixes: a1bdb71a32da ("net/mlx5: fix crash in Rx")
CC: stable at dpdk.org
Reported-by: Liming Sun
Completion queue entry data uses network endian, to access them we should use
ntoh*().
Fixes: c305090bbaf8 ("net/mlx5: replace countdown with threshold for Tx
completions")
CC: stable at dpdk.org
Reported-by: Liming Sun
Signed-off-by: Nelio Laranjeiro
---
drivers/net/mlx5/mlx5_rxtx.c | 2 +-
We can leave the title completion queue entry untouched since its contents
are not modified.
Reported-by: Liming Sun
Signed-off-by: Nelio Laranjeiro
---
drivers/net/mlx5/mlx5_rxtx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/mlx5/mlx5_rxtx.c
On Thu, Nov 17, 2016 at 01:49:25AM +0530, Jerin Jacob wrote:
> On Wed, Nov 16, 2016 at 06:00:00PM +, Harry van Haaren wrote:
> > This series of RFC patches implements the libeventdev API and a software
> > eventdev PMD.
> >
> > The implementation here is intended to enable the community to
On Thu, Nov 17, 2016 at 10:49:54AM +0100, Nelio Laranjeiro wrote:
> The list of segments to free was wrongly manipulated ending by only freeing
> the first segment instead of freeing all of them. The last one still
> belongs to the NIC and thus should not be freed.
>
> Fixes: a1bdb71a32da
On Thu, Nov 17, 2016 at 10:49:55AM +0100, Nelio Laranjeiro wrote:
> Completion queue entry data uses network endian, to access them we should use
> ntoh*().
>
> Fixes: c305090bbaf8 ("net/mlx5: replace countdown with threshold for Tx
> completions")
>
> CC: stable at dpdk.org
> Reported-by:
On Thu, Nov 17, 2016 at 10:49:56AM +0100, Nelio Laranjeiro wrote:
> We can leave the title completion queue entry untouched since its contents
> are not modified.
>
> Reported-by: Liming Sun
> Signed-off-by: Nelio Laranjeiro
> ---
> drivers/net/mlx5/mlx5_rxtx.c | 2 +-
> 1 file changed, 1
On Thu, Nov 17, 2016 at 6:30 AM, Shreyansh Jain
wrote:
> A device is connected to a bus and services by a driver associated with
> the bus. It is responsibility of the bus to identify the devices (scan)
> and then assign each device to a matching driver.
>
> A PMD would allocate a rte_xxx_driver
On Thu, Nov 17, 2016 at 6:29 AM, Shreyansh Jain
wrote:
> DPDK has been inherently a PCI inclined framework. Because of this, the
> design of device tree (or list) within DPDK is also PCI inclined. A non-PCI
> device doesn't have a way of being expressed without using hooks started from
> EAL to
On Thu, Nov 17, 2016 at 6:30 AM, Shreyansh Jain
wrote:
> From: Jan Viktorin
>
> Signed-off-by: Jan Viktorin
> Signed-off-by: Shreyansh Jain
> ---
> lib/librte_eal/common/include/rte_common.h | 18 ++
> 1 file changed, 18 insertions(+)
>
> diff --git
On Thu, Nov 17, 2016 at 6:30 AM, Shreyansh Jain
wrote:
> This patch demonstrates how eth_driver can be replaced with appropriate
> changes for rte_xxx_driver from the PMD itself. It uses ixgbe_ethernet as
> an example.
>
> A large set of changes exists in the rte_ethdev.c - primarily because too
Hello Jan,
Thanks for comments. Replies inline.
On Thursday 17 November 2016 04:49 PM, Jan Blunck wrote:
> On Thu, Nov 17, 2016 at 6:30 AM, Shreyansh Jain
> wrote:
>> A device is connected to a bus and services by a driver associated with
>> the bus. It is responsibility of the bus to identify
On Thursday 17 November 2016 05:36 PM, Jan Blunck wrote:
> On Thu, Nov 17, 2016 at 6:30 AM, Shreyansh Jain
> wrote:
>> From: Jan Viktorin
>>
>> Signed-off-by: Jan Viktorin
>> Signed-off-by: Shreyansh Jain
>> ---
>> lib/librte_eal/common/include/rte_common.h | 18 ++
>> 1 file
On Thursday 17 November 2016 05:25 PM, Jan Blunck wrote:
> On Thu, Nov 17, 2016 at 6:29 AM, Shreyansh Jain
> wrote:
>> DPDK has been inherently a PCI inclined framework. Because of this, the
>> design of device tree (or list) within DPDK is also PCI inclined. A non-PCI
>> device doesn't have a
Signed-off-by: Ferruh Yigit
---
doc/guides/prog_guide/kernel_nic_interface.rst | 2 ++
doc/guides/rel_notes/deprecation.rst | 6 ++
2 files changed, 8 insertions(+)
diff --git a/doc/guides/prog_guide/kernel_nic_interface.rst
b/doc/guides/prog_guide/kernel_nic_interface.rst
index
2016-11-17 17:42, Qiming Yang:
> This patch added API for 'rte_eth_dev_fwver_get'
>
> void rte_eth_dev_fwver_get(uint8_t port_id,
> char *fw_version, int fw_length);
Copying some code here doesn't help really help.
Could you describe what we can expect in this string?
How can we compare this
Lets clear the eth_dev->data when allocating a new rte_eth_dev so that
drivers only need to set non-zero values.
Signed-off-by: Jan Blunck
---
drivers/net/mlx4/mlx4.c | 1 -
drivers/net/mlx5/mlx5.c | 1 -
lib/librte_ether/rte_ethdev.c | 2 +-
3 files changed, 1 insertion(+), 3
This moves the non-PCI related initialization of the link state interrupt
callback list and the setting of the default MTU to rte_eth_dev_allocate()
so that drivers only need to set non-default values.
Signed-off-by: Jan Blunck
---
drivers/net/bonding/rte_eth_bond_api.c | 2 --
+Billy McFall
On 11/17/2016 04:49 AM, Yuanhan Liu wrote:
> On Thu, Nov 17, 2016 at 09:47:09AM +0100, Maxime Coquelin wrote:
>>
>> On 11/17/2016 09:29 AM, Yuanhan Liu wrote:
>>> As usaual, sorry for late response :/
>>>
>>> On Thu, Oct 13, 2016 at 08:50:52PM +0300, Michael S. Tsirkin wrote:
On 11/17/2016 2:24 PM, Jan Blunck wrote:
> This moves the non-PCI related initialization of the link state interrupt
> callback list and the setting of the default MTU to rte_eth_dev_allocate()
> so that drivers only need to set non-default values.
>
> Signed-off-by: Jan Blunck
> ---
>
2016-11-17 03:51, Xu, Qian Q:
> It is good that RC1 on Jan.11th is a hard deadline. We also need ensure that
> RC1 is a complete package with all features. If RC1 is out but missing some
> big features, then the test results on RC1 vs RC2 may have big difference.
> So, could we ensure that RC1
On 11/17/2016 2:24 PM, Jan Blunck wrote:
> Lets clear the eth_dev->data when allocating a new rte_eth_dev so that
> drivers only need to set non-zero values.
>
> Signed-off-by: Jan Blunck
> ---
> drivers/net/mlx4/mlx4.c | 1 -
> drivers/net/mlx5/mlx5.c | 1 -
>
On Thu, Nov 17, 2016 at 2:00 PM, Shreyansh Jain
wrote:
>
> On Thursday 17 November 2016 04:49 PM, Jan Blunck wrote:
>>
>> On Thu, Nov 17, 2016 at 6:30 AM, Shreyansh Jain
>> wrote:
>>>
>>> A device is connected to a bus and services by a driver associated with
>>> the bus. It is responsibility
On Thu, Nov 17, 2016 at 4:46 PM, Ferruh Yigit wrote:
> On 11/17/2016 2:24 PM, Jan Blunck wrote:
>> This moves the non-PCI related initialization of the link state interrupt
>> callback list and the setting of the default MTU to rte_eth_dev_allocate()
>> so that drivers only need to set
On Thu, Nov 17, 2016 at 2:08 PM, Shreyansh Jain
wrote:
> On Thursday 17 November 2016 05:25 PM, Jan Blunck wrote:
>>
>> On Thu, Nov 17, 2016 at 6:29 AM, Shreyansh Jain
>> wrote:
>>>
>>> DPDK has been inherently a PCI inclined framework. Because of this, the
>>> design of device tree (or list)
Lets clear the eth_dev->data when allocating a new rte_eth_dev so that
drivers only need to set non-zero values.
Signed-off-by: Jan Blunck
---
drivers/net/bonding/rte_eth_bond_api.c | 7 ---
drivers/net/mlx4/mlx4.c| 1 -
drivers/net/mlx5/mlx5.c| 1 -
This moves the non-PCI related initialization of the link state interrupt
callback list and the setting of the default MTU to rte_eth_dev_allocate()
so that drivers only need to set non-default values.
Signed-off-by: Jan Blunck
---
app/test/virtual_pmd.c | 2 --
In case of link speed set to 1Gb at peer side (with autoneg or with
defined speed) and cable not plugged-in when device is configured
and started, then link status is not updated properly with new speed
as no link setup is triggered.
To avoid this issue, IXGBE_FLAG_NEED_LINK_CONFIG is set to try
The cryptodev API had specified that if the digest address field was
left empty on an authentication operation, then the PMD would assume
the digest was appended to the source or destination data.
This case was not handled at all by most PMDs and incorrectly handled
by the QAT PMD.
As no bugs were
The cryptodev API had specified that if the digest address field was
left empty on an authentication operation, then the PMD would assume
the digest was appended to the source or destination data.
This case was not handled at all by most PMDs and incorrectly handled
by the QAT PMD.
As no bugs were
On Thu, Nov 17, 2016 at 05:49:36PM +0800, Yuanhan Liu wrote:
> On Thu, Nov 17, 2016 at 09:47:09AM +0100, Maxime Coquelin wrote:
> >
> >
> > On 11/17/2016 09:29 AM, Yuanhan Liu wrote:
> > >As usaual, sorry for late response :/
> > >
> > >On Thu, Oct 13, 2016 at 08:50:52PM +0300, Michael S.
On 11/17/2016 5:16 PM, Jan Blunck wrote:
> Lets clear the eth_dev->data when allocating a new rte_eth_dev so that
> drivers only need to set non-zero values.
>
> Signed-off-by: Jan Blunck
> ---
Reviewed-by: Ferruh Yigit
On 11/17/2016 5:16 PM, Jan Blunck wrote:
> This moves the non-PCI related initialization of the link state interrupt
> callback list and the setting of the default MTU to rte_eth_dev_allocate()
> so that drivers only need to set non-default values.
>
> Signed-off-by: Jan Blunck
Reviewed-by:
On Thu, Nov 17, 2016 at 09:20:50AM +, Mcnamara, John wrote:
> One committer to master represents a single point of failure and at times can
> be inefficient.
I have a lot more issues because of slow or inconclusive review of patches
than I do because of committers. Often times they just get
57 matches
Mail list logo