2016-07-21 11:07, Olivier Matz:
> On 07/19/2016 06:53 PM, Thomas Monjalon wrote:
> > The unavailable tests are filtered out by autotest by looking for
> > the symbols in the binary:
> >
> > PCI autotest: Skipped [Not Available] [00m 00s]
> > Malloc autotest: Su
2016-07-21 15:35, Thomas Monjalon:
> 2016-07-21 15:28, Olivier Matz:
> > Hi Santosh,
> >
> > On 07/21/2016 01:49 PM, Santosh Shukla wrote:
> > > test_mempool func not using pointer variable 'mp_ext' and incorrectly
> > > freed. So
> > > removing ptr var. Now freeing mp_stack var.
> > >
> > > Sig
> > rte_ring_create expects the size of the ring to
> > be a power of 2. REFCNT_RING_SIZE value is not
> > power of 2 in-case if RTE_MAX_LCORE == 96.
> > Fix it by aligning the size to next power of 2 value.
> >
> > Fixes: af75078f ("first public release")
> >
> > Signed-off-by: Jerin Jacob
>
>
> > Consumer queue dequeuing must be guaranteed to be done fully before the
> > tail is updated. This is not guaranteed with a read barrier,
> > changed to a write barrier just before tail update which in practice
> > guarantees correct order of reads and writes.
> >
> > Signed-off-by: Juhamatti
2016-07-21 16:25, Olivier Matz:
> On 07/21/2016 03:47 PM, Zoltan Kiss wrote:
> > On 21/07/16 14:40, Olivier Matz wrote:
> >> On 07/20/2016 07:16 PM, Zoltan Kiss wrote:
> >>> A recent patch brought up an issue about the size of the 'name' fields:
> >>>
> >>> 85cf0079 mem: avoid memzone/mempool/ring
2016-07-20 18:14, Zoltan Kiss:
> The new mempool handler interface forgets to register the free() function
> of the ops. Introduced in this patch:
>
> Fixes: 449c49b93a6b ("mempool: support handler operations")
>
> Signed-off-by: Zoltan Kiss
> Acked-by: Olivier Matz
Applied, thanks
> > (strlen(name) == sizeof(mz->name) - 1) is a valid case, change the
> > condition to reflect that.
> > Move it earlier to avoid lookup with invalid name.
> > Change errno to ENAMETOOLONG.
> >
> > Fixes: 85cf0079 ("mem: avoid memzone/mempool/ring name truncation")
> >
> > Signed-off-by: Zoltan
>
> This is an ABI deprecation notice for DPDK 16.11 in librte_ether about
> changes in rte_eth_dev and rte_eth_desc_lim structures.
>
> As discussed in that thread:
>
> http://dpdk.org/ml/archives/dev/2015-September/023603.html
>
> Different NIC models depending on HW offload requested might
2016-07-21 13:20, Jay Rolette:
> On Thu, Jul 21, 2016 at 10:33 AM, Ferruh Yigit
> wrote:
> > KNI ethtool is functional and maintained, and it may have users!
> >
> > Why just removing it, specially without providing an alternative?
> > Is is good time to discuss KCP again?
>
> Yes, my product use
On Thu, Jul 21, 2016 at 04:43:25PM +0300, Ilya Maximets wrote:
>
>
> On 21.07.2016 16:35, Yuanhan Liu wrote:
> > On Thu, Jul 21, 2016 at 04:19:35PM +0300, Ilya Maximets wrote:
> >> If something abnormal happened to QEMU, 'connect()' can block calling
> >> thread (e.g. main thread of OVS) forever
On Thu, Jul 21, 2016 at 04:19:35PM +0300, Ilya Maximets wrote:
> If something abnormal happened to QEMU, 'connect()' can block calling
> thread (e.g. main thread of OVS) forever or for a really long time.
> This can break whole application or block the reconnection thread.
>
> Example with OVS:
>
Hi Jerin,
Sorry, looks like I missed your reply. Please see below.
On Mon, Jul 11, 2016 at 04:11:43PM +0530, Jerin Jacob wrote:
> On Tue, Jul 05, 2016 at 08:16:46PM +0200, Adrien Mazarguil wrote:
>
> Hi Adrien,
>
> Overall this proposal looks very good. I could easily map to the
> classificatio
On Thu, Jul 21, 2016 at 03:55:36PM +0300, Ilya Maximets wrote:
> Currently while calling of 'rte_vhost_driver_unregister()' connection
> to QEMU will not be closed. This leads to inability to register driver
> again and reconnect to same virtual machine.
>
> This scenario is reproducible with OVS.
On Thu, Jul 21, 2016 at 03:58:11PM +0300, Ilya Maximets wrote:
> On 21.07.2016 15:58, Yuanhan Liu wrote:
> > On Thu, Jul 21, 2016 at 03:42:54PM +0300, Ilya Maximets wrote:
> >> On 21.07.2016 15:35, Yuanhan Liu wrote:
> >>> On Thu, Jul 21, 2016 at 03:13:14PM +0300, Ilya Maximets wrote:
> >>
> >>
On Thu, Jul 21, 2016 at 03:42:54PM +0300, Ilya Maximets wrote:
> On 21.07.2016 15:35, Yuanhan Liu wrote:
> > On Thu, Jul 21, 2016 at 03:13:14PM +0300, Ilya Maximets wrote:
>
> What do you think of it?
> >>>
> >>> I found that we can't check connection status without select/poll
> >>> on i
> -Original Message-
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Thursday, July 21, 2016 4:36 PM
> To: Jastrzebski, MichalX K
> Cc: dev at dpdk.org; Richardson, Bruce ;
> Kobylinski, MichalX ; Gonzalez Monroy,
> Sergio ; david.marchand at 6wind.com
> Subject: Re:
On Thu, Jul 21, 2016 at 02:03:38PM +0800, Wei Dai wrote:
Hi Wei,
> The '-' in lcores set overrides cpu set of following
> lcore set in the argument of EAL option --lcores.
>
> Fixes: 53e54bf81700 ("eal: new option --lcores for cpu assignment")
>
> Signed-off-by: Wei Dai
> ---
> lib/librte_eal/c
On Thu, Jul 21, 2016 at 03:13:14PM +0300, Ilya Maximets wrote:
> >>
> >> What do you think of it?
> >
> > I found that we can't check connection status without select/poll
> > on it. 'getsockopt()' will return 0 with no errors if connection
> > is not still established just like if it was.
> > So,
On Thu, Jul 21, 2016 at 02:14:59PM +0300, Ilya Maximets wrote:
> > Hmm, how about this fixup:
> > --
> > diff --git a/lib/librte_vhost/vhost_user/vhost-net-user.c
> > b/lib/librte_vhost/vhost_user/vhost-net-user.c
> > inde
Hi Robert,
Thank you for reviewing.
In the following case, the skip list is broken.
- Timer A and timer B are configured on the same lcore, in the same
pending list.
- The expire time of timer A is earlier than that of timer B.
- rte_timer_manage() is called on the lcore after the expire time of
81xx NIC subsystem differs in new PCI subsystem_device_id and
NICVF_CAP_CQE_RX2 capability.
Signed-off-by: Jerin Jacob
---
doc/guides/nics/thunderx.rst | 1 +
drivers/net/thunderx/base/nicvf_hw.c | 3 +++
drivers/net/thunderx/base/nicvf_hw.h | 1 +
drivers/net/thunderx/nicvf_ethdev.c |
Certain thunderx SoC pass has additional optional word
in Rx descriptor to hold tunneling extension info.
Based on this capability, the location where packet pointer
address stored in Rx descriptor will vary.
Signed-off-by: Jerin Jacob
---
drivers/net/thunderx/base/nicvf_hw.h | 5 +++--
drivers/
thunderx pmd driver needs to support multiple SoC
variants in ThunderX family.
Remove generic pass references from driver as each SoC
can have same pass number.
Signed-off-by: Jerin Jacob
---
drivers/net/thunderx/base/nicvf_hw.c | 2 +-
drivers/net/thunderx/base/nicvf_hw.h | 12 +---
dr
CN81xx is four core version of ThunderX SoC.
Added the support by adding new HW capability flag to select
the difference in runtime.
Jerin Jacob (3):
net/thunderx: remove generic passx references from the driver
net/thunderx: introduce cqe_rx2 HW capability flag
net/thunderx: add 81xx SoC s
On Thu, Jul 21, 2016 at 02:03:38PM +0300, Ilya Maximets wrote:
> From: Sergey Dyasly
>
> Rx loop inside _recv_raw_pkts_vec() ignores nb_pkts argument and always
> tries to receive RTE_I40E_VPMD_RX_BURST (32) packets. This is a violation
> of rte_eth_rx_burst() API and can lead to memory corruptio
Hi Rahul,
Please see below.
On Thu, Jul 21, 2016 at 01:43:37PM +0530, Rahul Lakkireddy wrote:
> Hi Adrien,
>
> The proposal looks very good. It satisfies most of the features
> supported by Chelsio NICs. We are looking for suggestions on exposing
> more additional features supported by Chelsio
On Wed, Jul 20, 2016 at 06:11:16PM +0100, Zoltan Kiss wrote:
Hi, Kiss
> Using weak symbols have a few issues with static linking:
>
> - normally the linker searches the .o files already linked, if your weak
> one is there, it won't check if there is a strong version
> - unless the symbol is dire
Dear Don,
On 07/21/2016 06:26 PM, Don Provan wrote:
>> -Original Message-
>> From: Ananyev, Konstantin [mailto:konstantin.ananyev at intel.com]
>> Sent: Thursday, July 21, 2016 3:51 AM
>> Subject: Re: [dpdk-dev] [PATCH 04/12] mbuf: add function to calculate a
>> checksum
>>
>> ...
>>> + A
2016-07-21 16:41, Igor Ryzhov:
> On Thu, Jul 21, 2016 at 4:33 PM, Ferruh Yigit
> wrote:
> > On 7/20/2016 5:07 PM, Thomas Monjalon wrote:
> > > The out-of-tree kernel code must be avoided.
> > > Moreover there is no good reason to keep this legacy feature
> > > which is only partially supported.
>
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 20 ++--
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 552a1de..55ff2c0 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/dr
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 63 -
1 file changed, 31 insertions(+), 32 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index f3df372..552a1de 100644
--- a/drivers/net/pcap/rt
Missing pcap assignment may cause pcap opened again, and previous one
not closed at all.
Fixes: 1e38a7c66923 ("pcap: fix storage of name and type in queues")
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/pcap/rte_
simplify function rte_eth_from_pcaps_common by using interim variables.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 27 +--
1 file changed, 13 insertions(+), 14 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pca
data->name assigned twice.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 9cb9861..83b05e2 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/
Both fields are fields of same type of struct, one's size can't be bigger
than others.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 4
1 file changed, 4 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index fb170db..9cb9861 1
Remove hardcoded single interface values, make it more obvious.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 13 ++---
1 file changed, 6 insertions(+), 7 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 8e011ea..fb170
Reorder functions to be able to remove function declarations in .c file.
Function definitions not modified.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 129
1 file changed, 64 insertions(+), 65 deletions(-)
diff --git a/drivers/net/
Remove unused and sort remaining.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 15 +++
1 file changed, 7 insertions(+), 8 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 3d85a7f..a142e38 100644
--- a/drivers/net/
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index ec25912..3d85a7f 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/pcap/rte_eth_
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 72 +
1 file changed, 37 insertions(+), 35 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 31eed58..ec25912 100644
--- a/drivers/net/pcap/rt
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 10 --
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 0445c74..31eed58 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/pc
Remove duplicated check by reorganizing the code, no functional change.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 49 ++---
1 file changed, 22 insertions(+), 27 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pc
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 3eacb82..7524293 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/pcap/r
Don't carry kvlist argument into sub function and used it, use kvlist
argument in upper level of call stack.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 30 +++---
1 file changed, 15 insertions(+), 15 deletions(-)
diff --git a/drivers/net/pcap/rte_e
Instead of defining numa_node variable upper level of call stack and
carry into sub function, set it where needs to be used.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 40 +---
1 file changed, 17 insertions(+), 23 deletions(-)
diff --gi
rte_eth_from_pcaps and rte_eth_from_pcaps_n_dumpers functions are very
close, updated rte_eth_from_pcaps function and reused.
No functional update.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 58 +
1 file changed, 13 insertions(+), 4
Number of queues defined by devargs, a check added to be sure this
number is not bigger than configured queue length.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 23 ---
1 file changed, 20 insertions(+), 3 deletions(-)
diff --git a/drivers/net/pcap/rte_
struct rx_pcaps and tx_pcaps used to point parsed devargs, but it is not
clear with current names.
Merged both into single struct and modified struct name and field names.
Functionality not changed, only struct names.
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 123 ++
Signed-off-by: Ferruh Yigit
---
drivers/net/pcap/rte_eth_pcap.c | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/drivers/net/pcap/rte_eth_pcap.c b/drivers/net/pcap/rte_eth_pcap.c
index 6343c0e..877e4c2 100644
--- a/drivers/net/pcap/rte_eth_pcap.c
+++ b/drivers/net/pc
pcap pmd is using ring pmd configuration parameters to set max number of
queues. This creates an unnecessary dependency and confusion.
Create a new config parameter for pcap PMD:
CONFIG_RTE_PMD_PCAP_MAX_QUEUES
Default value of config is same as ring parameter default.
pcap pmd doesn't need to be
No new feature added, code refactored.
This patch targetted for 16.11 release.
Ferruh Yigit (21):
net/pcap: create own configuration parameter
net/pcap: use macros for param string
net/pcap: reorganize private structs
net/pcap: add checks for max queue number
net/pcap: update function to
On Thu, Jul 21, 2016 at 12:45:32PM +0300, Ilya Maximets wrote:
> On 21.07.2016 12:37, Yuanhan Liu wrote:
> > On Thu, Jul 21, 2016 at 11:21:15AM +0300, Ilya Maximets wrote:
> >> If something abnormal happened to QEMU, 'connect()' can block calling
> >> thread (e.g. main thread of OVS) forever or for
On Thu, Jul 21, 2016 at 11:21:15AM +0300, Ilya Maximets wrote:
> If something abnormal happened to QEMU, 'connect()' can block calling
> thread (e.g. main thread of OVS) forever or for a really long time.
> This can break whole application or block the reconnection thread.
>
> Example with OVS:
>
This is an ABI deprecation notice for DPDK 16.11 in librte_ether about
changes in rte_eth_dev and rte_eth_desc_lim structures.
As discussed in that thread:
http://dpdk.org/ml/archives/dev/2015-September/023603.html
Different NIC models depending on HW offload requested might impose
different req
test_mempool func not using pointer variable 'mp_ext' and incorrectly freed. So
removing ptr var. Now freeing mp_stack var.
Signed-off-by: Santosh Shukla
---
app/test/test_mempool.c |3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/app/test/test_mempool.c b/app/test/test_m
On 21.07.2016 16:35, Yuanhan Liu wrote:
> On Thu, Jul 21, 2016 at 04:19:35PM +0300, Ilya Maximets wrote:
>> If something abnormal happened to QEMU, 'connect()' can block calling
>> thread (e.g. main thread of OVS) forever or for a really long time.
>> This can break whole application or block the
Hello.
I think good alternative is rte_ethtool library from ethtool sample
application.
But I am wondering why this code is only in app, not in lib.
Best regards,
Igor
On Thu, Jul 21, 2016 at 4:33 PM, Ferruh Yigit
wrote:
> On 7/20/2016 5:07 PM, Thomas Monjalon wrote:
> > The out-of-tree kernel
Hi,
2016-07-20 16:24, Michal Jastrzebski:
> - if (read(fd, &page, sizeof(uint64_t)) < 0) {
> +
> + retval = read(fd, &page, sizeof(uint64_t));
> + if (retval < 0) {
> RTE_LOG(ERR, EAL, "%s(): cannot read /proc/self/pagemap: %s\n",
> __func__,
On 7/20/2016 5:07 PM, Thomas Monjalon wrote:
> The out-of-tree kernel code must be avoided.
> Moreover there is no good reason to keep this legacy feature
> which is only partially supported.
>
> As described earlier in this plan:
> http://dpdk.org/ml/archives/dev/2016-July/043606.html
> it
On Tue, Jul 19, 2016 at 02:31:59PM +0200, Olivier Matz wrote:
> The support of virtio-user changed the way the mbuf dma address is
> retrieved, using a physical address in case of virtio-pci and a virtual
> address in case of virtio-user.
>
> This change introduced some possible memory corruption
2016-07-18 09:56, Ido Barnea:
> Signed-off-by: Ido Barnea
Patch reconstructed (format was not applicable),
and applied, thanks
> -Original Message-
> From: Ananyev, Konstantin [mailto:konstantin.ananyev at intel.com]
> Sent: Thursday, July 21, 2016 3:51 AM
> Subject: Re: [dpdk-dev] [PATCH 04/12] mbuf: add function to calculate a
> checksum
>
>...
> > + Added a new function ``rte_pktmbuf_cksum()`` to process the c
On 07/21/2016 03:47 PM, Zoltan Kiss wrote:
>
>
> On 21/07/16 14:40, Olivier Matz wrote:
>> Hi Zoltan,
>>
>>
>> On 07/20/2016 07:16 PM, Zoltan Kiss wrote:
>>> A recent patch brought up an issue about the size of the 'name' fields:
>>>
>>> 85cf0079 mem: avoid memzone/mempool/ring name truncation
Sometimes app just wants to update the RSS hash function and no RSS key
update is needed, but fm10k pmd will return EINVAL for this case.
If the rss_key is NULL, we don't need to check the rss_key_len.
Fixes: 57033cdf8fdc ("fm10k: add PF RSS")
Reported-by: Xueqin Lin
Signed-off-by: Xiao Wang
-
2016-07-15 10:30, Chao Zhu:
> On weak memory order architecture like POWER, rte_smp_wmb/rte_smp_rmb
> need to use CPU instructions, not compiler barrier. This patch fixes
> this. Also, to improve performance on PPC64, use light weight sync
> instruction instead of sync instruction.
>
> Signed-off-
On Wed, Jul 20, 2016 at 11:32:43AM +0300, Ilya Maximets wrote:
> Currently while calling of 'rte_vhost_driver_unregister()' connection
> to QEMU will not be closed. This leads to inability to register driver
> again and reconnect to same virtual machine.
>
> This scenario is reproducible with OVS.
The usual device sequence is configure, queue setup and start.
Crypto device should be started before use.
Signed-off-by: Akhil Goyal
Signed-off-by: Hemant Agrawal
---
examples/ipsec-secgw/ipsec-secgw.c | 4
1 file changed, 4 insertions(+)
diff --git a/examples/ipsec-secgw/ipsec-secgw.c
The usual device sequence is configure, queue setup and start.
Crypto device should be started before use.
Signed-off-by: Akhil Goyal
Signed-off-by: Hemant Agrawal
---
examples/l2fwd-crypto/main.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/examples/l2fwd-crypto/main.c b/examples/
Hi,
2016-07-21 14:03, Wei Dai:
> The '-' in lcores set overrides cpu set of following
> lcore set in the argument of EAL option --lcores.
>
> Fixes: 53e54bf81700 ("eal: new option --lcores for cpu assignment")
>
> Signed-off-by: Wei Dai
Thanks for the catch!
> --- a/lib/librte_eal/common/eal_
If something abnormal happened to QEMU, 'connect()' can block calling
thread (e.g. main thread of OVS) forever or for a really long time.
This can break whole application or block the reconnection thread.
Example with OVS:
ovs_rcu(urcu2)|WARN|blocked 512000 ms waiting for main to quiesce
On 21.07.2016 15:58, Yuanhan Liu wrote:
> On Thu, Jul 21, 2016 at 03:42:54PM +0300, Ilya Maximets wrote:
>> On 21.07.2016 15:35, Yuanhan Liu wrote:
>>> On Thu, Jul 21, 2016 at 03:13:14PM +0300, Ilya Maximets wrote:
>>
>> What do you think of it?
>
> I found that we can't check conne
I've fixed leak of file descriptors in 'vhost_user_remove_reconnect()'
and sent v3.
On 21.07.2016 11:31, Ilya Maximets wrote:
> Currently while calling of 'rte_vhost_driver_unregister()' connection
> to QEMU will not be closed. This leads to inability to register driver
> again and reconnect to sa
Currently while calling of 'rte_vhost_driver_unregister()' connection
to QEMU will not be closed. This leads to inability to register driver
again and reconnect to same virtual machine.
This scenario is reproducible with OVS. While executing of the following
command vhost port will be re-created (
On Thu, Jul 21, 2016 at 3:32 PM, Thomas Monjalon
wrote:
> 2016-07-21 13:20, Jay Rolette:
> > On Thu, Jul 21, 2016 at 10:33 AM, Ferruh Yigit
> > wrote:
> > > KNI ethtool is functional and maintained, and it may have users!
> > >
> > > Why just removing it, specially without providing an alternati
Hi,
On 07/20/2016 07:16 PM, Zoltan Kiss wrote:
> (strlen(name) == sizeof(mz->name) - 1) is a valid case, change the
> condition to reflect that.
> Move it earlier to avoid lookup with invalid name.
> Change errno to ENAMETOOLONG.
>
> Fixes: 85cf0079 ("mem: avoid memzone/mempool/ring name truncati
On 21.07.2016 15:35, Yuanhan Liu wrote:
> On Thu, Jul 21, 2016 at 03:13:14PM +0300, Ilya Maximets wrote:
What do you think of it?
>>>
>>> I found that we can't check connection status without select/poll
>>> on it. 'getsockopt()' will return 0 with no errors if connection
>>> is not still
Hi Jerin,
On 07/18/2016 07:55 AM, Jerin Jacob wrote:
> rte_ring_create expects the size of the ring to
> be a power of 2. REFCNT_RING_SIZE value is not
> power of 2 in-case if RTE_MAX_LCORE == 96.
> Fix it by aligning the size to next power of 2 value.
>
> Fixes: af75078f ("first public release")
Hi Zoltan,
On 07/20/2016 07:16 PM, Zoltan Kiss wrote:
> A recent patch brought up an issue about the size of the 'name' fields:
>
> 85cf0079 mem: avoid memzone/mempool/ring name truncation
>
> These relations should be observed:
>
> 1. Each ring creates a memzone with a prefixed name:
> RTE_RI
On 20/07/2016 15:24, Michal Jastrzebski wrote:
> In rte_mem_virt2phy: Value returned from a function and indicating the
> number of bytes was ignored. This could cause a wrong pfn (page frame
> number) mask read from pagemap file.
> When read returns less than the number of sizeof(uint64_t) bytes,
Hi Sugesh,
I do not have much to add, please see below.
On Thu, Jul 21, 2016 at 11:06:52AM +, Chandran, Sugesh wrote:
[...]
> > > RSS hashing support :- Just to confirm, the RSS flow action allows
> > > application to decide the header fields to produce the hash. This
> > > gives programmabil
2016-07-21 15:28, Olivier Matz:
> Hi Santosh,
>
> On 07/21/2016 01:49 PM, Santosh Shukla wrote:
> > test_mempool func not using pointer variable 'mp_ext' and incorrectly
> > freed. So
> > removing ptr var. Now freeing mp_stack var.
> >
> > Signed-off-by: Santosh Shukla
> > ---
> > app/test/tes
Hi Santosh,
On 07/21/2016 01:49 PM, Santosh Shukla wrote:
> test_mempool func not using pointer variable 'mp_ext' and incorrectly freed.
> So
> removing ptr var. Now freeing mp_stack var.
>
> Signed-off-by: Santosh Shukla
> ---
> app/test/test_mempool.c |3 +--
> 1 file changed, 1 insertio
> On Jul 21, 2016, at 10:06 AM, Neil Horman wrote:
>
> On Thu, Jul 21, 2016 at 02:09:19PM +, Wiles, Keith wrote:
>>
>>> On Jul 21, 2016, at 8:54 AM, Neil Horman wrote:
>>>
>>> On Wed, Jul 20, 2016 at 10:32:28PM +, Wiles, Keith wrote:
> On Jul 20, 2016, at 3:16 PM, Neil Horma
On 21.07.2016 15:10, Ilya Maximets wrote:
> On 21.07.2016 14:40, Yuanhan Liu wrote:
>> On Thu, Jul 21, 2016 at 02:14:59PM +0300, Ilya Maximets wrote:
Hmm, how about this fixup:
--
diff --git a/lib/librte_vh
On 21.07.2016 14:40, Yuanhan Liu wrote:
> On Thu, Jul 21, 2016 at 02:14:59PM +0300, Ilya Maximets wrote:
>>> Hmm, how about this fixup:
>>> --
>>> diff --git a/lib/librte_vhost/vhost_user/vhost-net-user.c
>>> b/lib/librte_
2016-07-21 12:01, Mrozowicz, SlawomirX:
> Hi Thomas,
>
> As I understand Sergio suggested to come back to the solution similar to v1.
> Could you comment or better take decision which solution should be applied,
> please.
>
> Best Regards,
> S?awomir
>
>
> >-Original Message-
> >From:
On 21/07/16 14:40, Olivier Matz wrote:
> Hi Zoltan,
>
>
> On 07/20/2016 07:16 PM, Zoltan Kiss wrote:
>> A recent patch brought up an issue about the size of the 'name' fields:
>>
>> 85cf0079 mem: avoid memzone/mempool/ring name truncation
>>
>> These relations should be observed:
>>
>> 1. Each ri
Hi Wenzhuo,
It seems that we agree on about everything now, just a few more comments
below after snipping the now irrelevant parts.
On Thu, Jul 21, 2016 at 03:18:11AM +, Lu, Wenzhuo wrote:
[...]
> > > > > Does it mean PMD should store and maintain all the rules? Why not
> > > > > let rte do
>
On Thu, Jul 21, 2016 at 03:22:45PM +, Wiles, Keith wrote:
>
> > On Jul 21, 2016, at 10:06 AM, Neil Horman wrote:
> >
> > On Thu, Jul 21, 2016 at 02:09:19PM +, Wiles, Keith wrote:
> >>
> >>> On Jul 21, 2016, at 8:54 AM, Neil Horman wrote:
> >>>
> >>> On Wed, Jul 20, 2016 at 10:32:28PM
On Thu, 21 Jul 2016 10:08:23 +0200
Olivier Matz wrote:
> +/**
> + * Deprecated.
> + * Checking this flag alone is deprecated: check the 2 bits of
> + * PKT_RX_L4_CKSUM_MASK.
> + * This flag was set when the L4 checksum of a packet was detected as
> + * wrong by the hardware.
> + */
> +#define PKT
On Thu, 21 Jul 2016 10:08:20 +0200
Olivier Matz wrote:
> + dev_info->max_rx_queues = (uint16_t)
> + ((VIRTIO_MAX_RX_QUEUES < hw->max_queue_pairs) ?
> + VIRTIO_MAX_RX_QUEUES : hw->max_queue_pairs);
> + dev_info->max_tx_queues = (uint16_t)
> + ((V
On 21.07.2016 13:37, Ilya Maximets wrote:
>
>
> On 21.07.2016 13:13, Yuanhan Liu wrote:
>> On Thu, Jul 21, 2016 at 12:45:32PM +0300, Ilya Maximets wrote:
>>> On 21.07.2016 12:37, Yuanhan Liu wrote:
On Thu, Jul 21, 2016 at 11:21:15AM +0300, Ilya Maximets wrote:
> If something abnormal hap
If something abnormal happened to QEMU, 'connect()' can block calling
thread (e.g. main thread of OVS) forever or for a really long time.
This can break whole application or block the reconnection thread.
Example with OVS:
ovs_rcu(urcu2)|WARN|blocked 512000 ms waiting for main to quiesce
- Fix vhost setup flags
- Add minor edits to improve readability and consistency
---
v2: - revert file mode changes made erroneously in v1
Signed-off-by: Mark Kavanagh
---
doc/guides/sample_app_ug/tep_termination.rst | 8
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/d
> On Jul 21, 2016, at 8:54 AM, Neil Horman wrote:
>
> On Wed, Jul 20, 2016 at 10:32:28PM +, Wiles, Keith wrote:
>>
>>> On Jul 20, 2016, at 3:16 PM, Neil Horman wrote:
>>>
>>> On Wed, Jul 20, 2016 at 07:47:32PM +, Wiles, Keith wrote:
> On Jul 20, 2016, at 12:48 PM, Neil Horma
- Fix vhost setup flags
- Add minor edits to improve readability and consistency
---
v2: - revert file mode changes made erroneously in v1
Signed-off-by: Mark Kavanagh
---
doc/guides/sample_app_ug/tep_termination.rst | 8
1 file changed, 4 insertions(+), 4 deletions(-)
mode change 10
The '-' in lcores set overrides cpu set of following
lcore set in the argument of EAL option --lcores.
Fixes: 53e54bf81700 ("eal: new option --lcores for cpu assignment")
Signed-off-by: Wei Dai
---
lib/librte_eal/common/eal_common_options.c | 12
1 file changed, 12 insertions(+)
d
From: Sergey Dyasly
Rx loop inside _recv_raw_pkts_vec() ignores nb_pkts argument and always
tries to receive RTE_I40E_VPMD_RX_BURST (32) packets. This is a violation
of rte_eth_rx_burst() API and can lead to memory corruption (out-of-bounds
writes to struct rte_mbuf **rx_pkts) if nb_pkts is less
Hi Amin,
> -Original Message-
> From: Tootoonchian, Amin
> Sent: Wednesday, July 20, 2016 5:08 PM
> To: Kerlin, MarcinX
> Cc: dev at dpdk.org; thomas.monjalon at 6wind.com
> Subject: RE: [PATCH] ethdev: ensure consistent port id assignment
>
> Hi Marcin,
>
> Comments inline:
>
> >
Hi Adrien,
The proposal looks very good. It satisfies most of the features
supported by Chelsio NICs. We are looking for suggestions on exposing
more additional features supported by Chelsio NICs via this API.
Chelsio NICs have two regions in which filters can be placed -
Maskfull and Maskless
1 - 100 of 143 matches
Mail list logo
