2016-05-31 10:40, xu, huilong:
> maybe we should change log level, when add port in blacklist,for check it
> easy.
> and it not influence performance and function
Please, could you show an example of the device init logs at INFO level
before and after this patch?
Thanks
Hi,
Thank you very much for your replies. I removed this API, I also see
rte_eal_pci_set_blacklist() is no more available with DPDK 1.7.1, I am
using rte_eal_devargs_add(). Any document mentioning changes from DPDK-1.6
to DPDK-1.7?
Thanks,
Shiva
On Tue, May 24, 2016 at 11:32 AM, Wiles, Keith
w
Hi Ilya,
Generically speaking, this patch looks good to me. But I guess still
need more time to check this issue later; I still failed to reproduce
it on my side after all. So, please allow a late merge.
Thanks.
--yliu
On Mon, May 30, 2016 at 02:05:07PM +0300, Ilya Maximets wrote:
> Pin
2016-05-30 14:04, Olivier Matz:
> Enhance the logs related to mbuf pool creation. Display an info level
> log when creating the mbuf, and display the error as a string on failure.
>
> After the patch, we have:
>
> [...]
> EAL: probe driver: 8086:10fb rte_ixgbe_pmd
> USER1: create a new mb
Hi Ferruh,
On 05/30/2016 05:49 PM, Ferruh Yigit wrote:
> On 5/30/2016 12:56 PM, Olivier Matz wrote:
>> The $(comma) variable is not defined in this Makefile, nor in
>> any included Makefile. Seen while doing a "make clean" on ubuntu:
>>
>> $ make clean
>> == Clean lib
>> == Clean lib/librte_
This patch enables rss (receive side scaling) per network interface
through the configuration file. The user can specify following
parameters in LINK section for enabling the rss feature - rss_qs,
rss_proto_ipv4, rss_proto_ipv6 and ip_proto_l2.
The "rss_qs" is mandatory parameter which indicates t
On Mon, May 30, 2016 at 05:26:21PM +0200, Olivier Matz wrote:
> Hi,
>
> I'm planning to add the support for offloads in virtio-net pmd.
> It appears that the current rx flags in mbuf are not sufficient to
> describe the state of a packet received from a virtual driver.
> I think we need a way to s
On 5/30/2016 5:21 PM, Olivier Matz wrote:
> Hi Ferruh,
>
> On 05/30/2016 05:49 PM, Ferruh Yigit wrote:
>> On 5/30/2016 12:56 PM, Olivier Matz wrote:
>>> The $(comma) variable is not defined in this Makefile, nor in
>>> any included Makefile. Seen while doing a "make clean" on ubuntu:
>>>
>>> $ m
On Tue, May 10, 2016 at 10:51:59AM +0800, Jingjing Wu wrote:
> Previously, if message is sent successfully, but no response is
> received, function "i40evf_execute_vf_cmd" will return without error.
> The root cause is value "err" is overwritten. This patch fixes it.
>
> Fixes: ae19955e7c86 ("i40e
Hi,
I'm planning to add the support for offloads in virtio-net pmd.
It appears that the current rx flags in mbuf are not sufficient to
describe the state of a packet received from a virtual driver.
I think we need a way to say "the checksum in the packet data is
not calculated, but the integrity o
We keep a common vq structure, containing only vq related fields,
and then split others into RX, TX and control queue respectively.
Signed-off-by: Huawei Xie
---
v2:
- don't split virtio_dev_rx/tx_queue_setup
v3:
- fix some 80 char warnings
- fix other newer version checkpatch warnings
- remove '
On 5/30/2016 12:56 PM, Olivier Matz wrote:
> The $(comma) variable is not defined in this Makefile, nor in
> any included Makefile. Seen while doing a "make clean" on ubuntu:
>
> $ make clean
> == Clean lib
> == Clean lib/librte_compat
> == Clean lib/librte_eal
> == Clean lib/librte_eal/
From: Martinx - ? [mailto:thiagocmarti...@gmail.com]
Sent: Monday, May 30, 2016 5:01 PM
To: Bodireddy, Bhanuprakash
Cc: Christian Ehrhardt ; ; dev ; qemu-stable
at nongnu.org
Subject: Re: [ovs-dev] If 1 KVM Guest loads the virtio-pci, on top of
dpdkvhostuser OVS socket interface, it slows d
On Wed, May 25, 2016 at 12:16:41AM +0800, Huawei Xie wrote:
> There is no external function call or any barrier in the loop,
> the used->idx would only be retrieved once.
>
> Signed-off-by: Huawei Xie
> ---
> drivers/net/virtio/virtio_ethdev.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deleti
+;
I40E_WRITE_FLUSH(hw);
- if (pf->floating == true) {
+ if (pf->floating == true &&
+ (pf->vf_fbitmap & 1 << vf_id)) {
vf->vsi = i40e_vsi_setup(vf->pf, I40E_VSI_SRIOV,
NULL, vf->vf_idx);
} else {
--
2.1.4
-- next part --
An embedded and charset-unspecified text was scrubbed...
Name: floating-VEB-testplan - attach.txt
URL:
<http://dpdk.org/ml/archives/dev/attachments/20160530/8259c3de/attachment-0001.txt>
On 2016/05/27 21:28, Mcnamara, John wrote:
>> -Original Message-
>> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Sean Harte
>> Sent: Friday, May 27, 2016 1:27 PM
>> To: mukawa at igel.co.jp
>> Cc: dev at dpdk.org; Harte, Sean
>> Subject: [dpdk-dev] [PATCH v1] null: set port_id i
On Fri, May 06, 2016 at 09:21:30PM -0700, Rasesh Mody wrote:
> Rename RTE_LIBRTE_QEDE_DEBUG_DRV to RTE_LIBRTE_QEDE_DEBUG_DRIVER
>
> Fixes: 425cba2a5176 ("qede: enable PMD build")
> Fixes: 33e9ff1b72ca ("qede: add core driver")
>
These fixes lines don't have the correct commit id's in them. The co
On Fri, May 06, 2016 at 09:21:31PM -0700, Rasesh Mody wrote:
> From: Harish Patil
>
> Under certain scenarios, MFW periodically polls the driver
> for LAN statistics. This patch implements the osal hook to
> fill in the stats.
>
> Fixes: ffa002d318d36 ("qede: add base driver")
>
What is MFW?
/
This commit adds following changes to configuration file parsing of
the ip pipeline application;
1. Parsing routines related to packet queues (pktq_in/out fields in the
PIPELINE section) and message queues (msgq_in/out fields of in the MSGQ
Section) are updated.
In the parsing routines, function
Hi Arek,
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Arek Kusztal
> Sent: Monday, May 30, 2016 1:39 PM
> To: dev at dpdk.org
> Cc: Trahe, Fiona; Griffin, John; Jain, Deepak K; olivier.matz at 6wind.com;
> thomas.monjalon at 6wind.com; Kusztal, ArkadiuszX
On 30.05.2016 15:00, Tan, Jianfeng wrote:
> Hi,
>
>> -Original Message-
>> From: Ilya Maximets [mailto:i.maximets at samsung.com]
>> Sent: Friday, May 20, 2016 8:50 PM
>> To: dev at dpdk.org; Xie, Huawei; Yuanhan Liu
>> Cc: Dyasly Sergey; Heetae Ahn; Tan, Jianfeng; Ilya Maximets
>> Subject
On Fri, May 27, 2016 at 03:44:31PM +0100, Hunt, David wrote:
>
>
Hi David,
[snip]
> That chunk of code above would be better moved all right. I'd suggest
> moving it to the
> rte_mempool_create function, as that's the one that needs the backward
> compatibility.
OK
>
> On the flags issue,
-Original Message-
From: Kusztal, ArkadiuszX
Sent: Monday, May 30, 2016 1:39 PM
To: dev at dpdk.org
Cc: Trahe, Fiona ; Griffin, John ; Jain, Deepak K ; olivier.matz at
6wind.com; thomas.monjalon at 6wind.com; Kusztal, ArkadiuszX
Subject: [PATCH] qat: fix phys address of content descrip
Hi Wei Jie,
Thank you for submitting this patch. I am currently travelling, I will review
your patch and come back to you hopefully later this week or early next week.
Regards,
Cristian
> -Original Message-
> From: WeiJie Zhuang [mailto:zhuangwj at gmail.com]
> Sent: Saturday, May 28, 2
Ping.
Best regards, Ilya Maximets.
On 23.05.2016 14:04, Ilya Maximets wrote:
> On 23.05.2016 13:57, Yuanhan Liu wrote:
>> On Fri, May 20, 2016 at 03:50:04PM +0300, Ilya Maximets wrote:
>>> In current implementation guest application can reinitialize vrings
>>> by executing start after stop. In th
Enhance the logs related to mbuf pool creation. Display an info level
log when creating the mbuf, and display the error as a string on failure.
After the patch, we have:
[...]
EAL: probe driver: 8086:10fb rte_ixgbe_pmd
USER1: create a new mbuf pool : n=331456, \
size=2176, socket=0
On 05/30/2016 12:39 PM, zr at semihalf.com wrote:
> From: Zyta Szpak
>
> Version 2 of fixing the fixed register width assumption.
> rte_eth_dev_get_reg_length and rte_eth_dev_get_reg callbacks
> do not provide register size to the app in any way. It is
> needed to allocate proper number of bytes b
The $(comma) variable is not defined in this Makefile, nor in
any included Makefile. Seen while doing a "make clean" on ubuntu:
$ make clean
== Clean lib
== Clean lib/librte_compat
== Clean lib/librte_eal
== Clean lib/librte_eal/common
== Clean lib/librte_eal/linuxapp
== Clean lib/li
From: Arkadiusz Kusztal
this patch fixes an error with computation of physical address of
content descriptor in the symmetric operations session
Signed-off-by: Arkadiusz Kusztal
---
drivers/crypto/qat/qat_crypto.c | 9 ++---
lib/librte_cryptodev/rte_cryptodev.c | 2 +-
2 files changed
Answers inline, as follows:
On 30 May 2016 at 12:44, Bodireddy, Bhanuprakash <
bhanuprakash.bodireddy at intel.com> wrote:
> *From:* Martinx - ? [mailto:thiagocmartinsc at gmail.com]
> *Sent:* Monday, May 30, 2016 5:01 PM
> *To:* Bodireddy, Bhanuprakash
> *Cc:* Christian Ehrhardt ; <
> dev a
On 5/30/2016 10:41 AM, Jerin Jacob wrote:
--snip--
>> Of course, that won't help if we need to pass in more data, in which case
>> we'd probably need an
>> opaque data pointer somewhere. It would probably be most useful to pass it
>> in with the
>> alloc, which may need the data. Any suggestions?
Signed-off-by: Ferruh Yigit
---
lib/librte_eal/linuxapp/kni/compat.h | 9 +
lib/librte_eal/linuxapp/kni/kni_misc.c | 30 --
lib/librte_eal/linuxapp/kni/kni_net.c | 8
3 files changed, 29 insertions(+), 18 deletions(-)
diff --git a/lib/librte_eal/
Signed-off-by: Ferruh Yigit
---
lib/librte_eal/linuxapp/kni/compat.h | 12 +---
1 file changed, 5 insertions(+), 7 deletions(-)
diff --git a/lib/librte_eal/linuxapp/kni/compat.h
b/lib/librte_eal/linuxapp/kni/compat.h
index 0e939e4..d10040d 100644
--- a/lib/librte_eal/linuxapp/kni/compat
Fix compile error becuase of Linux API change, 'trans_start' field
removed from 'struct net_device'.
Linux: 9b36627acecd ("net: remove dev->trans_start")
Signed-off-by: Ferruh Yigit
---
lib/librte_eal/linuxapp/kni/compat.h | 4
lib/librte_eal/linuxapp/kni/kni_net.c | 9 -
2 files
Hello Bhanu,
I'm a little bit confused, you said that the problem can be fixed but,
later, you also said that:
"On a Multi VM setup even with the above patch applied, one might see
aggregate throughput difference when vNIC is bind to igb_uio vs
virtio-pci"...
My idea is to use OVS with DPDK i
Hi,
> -Original Message-
> From: Ilya Maximets [mailto:i.maximets at samsung.com]
> Sent: Friday, May 20, 2016 8:50 PM
> To: dev at dpdk.org; Xie, Huawei; Yuanhan Liu
> Cc: Dyasly Sergey; Heetae Ahn; Tan, Jianfeng; Ilya Maximets
> Subject: [PATCH] vhost: fix segfault on bad descriptor addr
Signed-off-by: Remy Horton
---
doc/guides/prog_guide/poll_mode_drv.rst | 25 +
1 file changed, 21 insertions(+), 4 deletions(-)
diff --git a/doc/guides/prog_guide/poll_mode_drv.rst
b/doc/guides/prog_guide/poll_mode_drv.rst
index 7698692..6cd86dd 100644
--- a/doc/guides/p
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch removes the name field
and all its usage of the old API.
Signed-off-by: Remy Horton
---
drivers/net/e10
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the proc_info
application to use the new API that seperates name string and value
queries.
Signed
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the test-pmd
application to use the new API that seperates name string and value
queries.
Signed-
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the virtio driver
to use the new API that seperates name string and value queries.
Signed-off-by:
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the i40e driver
to use the new API that seperates name string and value queries.
Signed-off-by: R
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the fm10k driver
to use the new API that seperates name string and value queries.
Signed-off-by:
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the e1000 driver
to use the new API that seperates name string and value queries.
Signed-off-by:
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the ixgbe driver
to use the new API that seperates name string and value queries.
Signed-off-by:
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patch changes the API for
xstats to use integer identifiers instead of strings.
Signed-off-by: Remy Horton
---
The current extended ethernet statistics fetching involve doing several
string operations, which causes performance issues if there are lots of
statistics and/or network interfaces. This patchset changes the API for
xstats to use integer identifiers instead of strings and implements
this new API fo
From: Zyta Szpak
Version 2 of fixing the fixed register width assumption.
Not every device uses 32-bit wide register. The app was allocating too
little space for 64-bit registers which resulted in memory corruption.
This commit resolves this by getting the size of register in bytes for
a specific
From: Zyta Szpak
Version 2 of fixing the fixed register width assumption.
rte_eth_dev_get_reg_length and rte_eth_dev_get_reg callbacks
do not provide register size to the app in any way. It is
needed to allocate proper number of bytes before retrieving
registers content with rte_eth_dev_get_reg.
From: Zyta Szpak
Version 2 of fixing the fixed register width assumption.
rte_eth_dev_get_reg_length and rte_eth_dev_get_reg callbacks
do not provide register size to the app in any way. It is
needed to allocate proper number of bytes before retrieving
registers content with rte_eth_dev_get_reg.
On 27.05.2016 12:28, Panu Matilainen wrote:
> On 05/25/2016 09:36 AM, zr at semihalf.com wrote:
>> From: Zyta Szpak
>>
>> Version 2 of fixing the fixed register width assumption.
>> rte_eth_dev_get_reg_length and rte_eth_dev_get_reg callbacks
>> do not provide register size to the app in any way
On Mon, May 30, 2016 at 02:40:00AM +, Xie, Huawei wrote:
> On 5/27/2016 5:06 PM, Yuanhan Liu wrote:
> > On Tue, May 24, 2016 at 09:38:32PM +0800, Huawei Xie wrote:
> >>vq->vq_ring_mem = mz->phys_addr;
> >>vq->vq_ring_virt_mem = mz->addr;
> >> - PMD_INIT_LOG(DEBUG, "vq->vq_ring_mem:
Hi,
It is the standard DPDK return value -ENOTSUP when the function is not
supported by Ethernet device. I think it is safer to keep it this way
rather than default implicitly to sizeof(uint32_t) and more generic.
Regards,
Zyta
On 25.05.2016 15:14, Remy Horton wrote:
> 'noon,
>
> Was expecting
Signed-off-by: Huawei Xie
Signed-off-by: Jianfeng Tan
---
doc/guides/rel_notes/release_16_07.rst | 4
1 file changed, 4 insertions(+)
diff --git a/doc/guides/rel_notes/release_16_07.rst
b/doc/guides/rel_notes/release_16_07.rst
index f6d543c..b1054b6 100644
--- a/doc/guides/rel_notes/relea
Add a new virtual device named vhost-user, which can be used just like
eth_ring, eth_null, etc. To reuse the code of original virtio, we do
some adjustment in virtio_ethdev.c, such as remove key _static_ of
eth_virtio_dev_init() so that it can be reused in virtual device; and
we add some check to m
This patch implements another new instance of struct virtio_pci_ops to
drive the virtio-user virtual device. Instead of rd/wr ioport or PCI
configuration space, this virtual pci driver will rd/wr the virtual
device struct virtio_user_hw, and when necessary, invokes APIs provided
by device emulation
Two device emulation layer APIs are added for virtio driver to call:
- virtio_user_start_device()
- virtio_user_stop_device()
These APIs will get called by virtio driver, and they call vhost adapter
layer APIs to implement the functionality. Besides, this patch defines
a struct named virtio_us
This patch is to provide vhost adapter layer implementations. Instead
of relying on a hypervisor to translate between device emulation and
vhost backend, here we directly talk with vhost backend through the
vhost file. Depending on the type of vhost file,
- vhost-user is used if the given path po
This patch is related to how to calculate relative address for vhost
backend.
The principle is that: based on one or multiple shared memory regions,
vhost maintains a reference system with the frontend start address,
backend start address, and length for each segment, so that each
frontend address
Abstract vring hdr desc init as an inline method.
Signed-off-by: Huawei Xie
Signed-off-by: Jianfeng Tan
---
drivers/net/virtio/virtio_ethdev.c | 42 ++
1 file changed, 24 insertions(+), 18 deletions(-)
diff --git a/drivers/net/virtio/virtio_ethdev.c
b/drive
This patch is to move phys addr check from virtio_dev_queue_setup
to pci ops. To makt that happen, make sure virtio_ops.setup_queue
return the result if we pass through the check.
Signed-off-by: Jianfeng Tan
Signed-off-by: Huawei Xie
Acked-by: Yuanhan Liu
---
drivers/net/virtio/virtio_ethdev.c
v5:
- Rename struct virtio_user_hw to struct virtio_user_dev.
- Rename "vdev_private" to "virtio_user_dev".
- Move special handling into virtio_ethdev.c from queue_setup().
- Add vring in virtio_user_dev (remove rte_eth_dev_data), so that
device does not depend on driver's data structure (rt
Hi David,
On 05/26/2016 04:15 PM, David Hunt wrote:
> recent CPU's can easily wrap around a 32-bit unsigned int in
> the mempool perf test. Increase to a 64-bit uint.
>
> v2: change from %lu to %"PRIu64"
>
> Signed-off-by: David Hunt
Acked-by: Olivier Matz
On 05/24/2016 07:15 AM, Yuanhan Liu wrote:
> On Mon, May 23, 2016 at 03:40:58PM +0200, Olivier Matz wrote:
>> For reference, here is the report of the ABI checker for EAL:
>>
>> [?] struct rte_pci_ioport (2)
>>
>> 1 Field len has been added to this type.
>>1) This field will not be initializ
Hi Jerin,
On 05/26/2016 10:07 AM, Jerin Jacob wrote:
> Signed-off-by: Jerin Jacob
> ---
> v1..v2
> Corrected the the git commit message(s/mbuf/mempool/g)
> ---
> lib/librte_mempool/rte_mempool.h | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/lib/librte_mempool/rte
On 05/27/2016 05:05 PM, Thomas Monjalon wrote:
> 2016-05-27 17:12, Jerin Jacob:
>> IMHO, I think we should have means to abstract this _logical_ changes
>> under conditional compilation flags and any arch/platform can choose
>> to select what it suites better for that arch/platform.
>>
>> We may
Hi,
> uint16_t enic_xmit_pkts(void *tx_queue, struct rte_mbuf **tx_pkts,
> uint16_t nb_pkts)
> {
> uint16_t index;
> - unsigned int frags;
> - unsigned int pkt_len;
> - unsigned int seg_len;
> - unsigned int inc_len;
> + unsigned int pkt_len, data_len;
> uns
On 5/30/2016 11:00 AM, Yuanhan Liu wrote:
> On Mon, May 30, 2016 at 02:40:00AM +, Xie, Huawei wrote:
>> On 5/27/2016 5:06 PM, Yuanhan Liu wrote:
>>> On Tue, May 24, 2016 at 09:38:32PM +0800, Huawei Xie wrote:
vq->vq_ring_mem = mz->phys_addr;
vq->vq_ring_virt_mem = mz->addr;
Tested-by: Peng Yuan
- Test Commit: a3f9ec846f9e7347d3a98da52256607345b4861d
- OS/Kernel: Fedora 23/4.2.3
- GCC: gcc version 5.3.1 20151207 (Red Hat 5.3.1-2) (GCC)
- CPU: Intel(R) Xeon(R) CPU E5-2699 v3 @ 2.30GHz
- Total 7 cases, 7 passed, 0 failed.
pf
vf(pf_kerneldriver)
mac_filter
On 5/27/2016 5:06 PM, Yuanhan Liu wrote:
> On Tue, May 24, 2016 at 09:38:32PM +0800, Huawei Xie wrote:
>> vq->vq_ring_mem = mz->phys_addr;
>> vq->vq_ring_virt_mem = mz->addr;
>> -PMD_INIT_LOG(DEBUG, "vq->vq_ring_mem: 0x%"PRIx64,
>> (uint64_t)mz->phys_addr);
>> -PMD_INIT_LOG(
70 matches
Mail list logo
