2014-10-31 15:37, Jincheng Miao:
> From upstream kernel commit 3db2e9cd, strict_strto* serial functions
> are removed. So that we should directly used kstrtoul instead.
>
> For compatible with old kernel and RHEL6, add some logic to
> igb_uio/compat.h, same as what we do for pci_num_vf().
>
>
2014-11-28 16:13, Jincheng Miao:
>
> On 11/28/2014 01:01 AM, Thomas Monjalon wrote:
> > 2014-10-31 15:37, Jincheng Miao:
> >> Function pci_num_vf() is introduced from upstream linux-2.6.34. So
> >> this patch make compatible with longterm kernel linux-2.6.32.63.
> >>
> >> For RHEL6's kernel,
> > Use of \n in sed expression is not portable and triggered an invalid
> > configuration on BSD (at least).
> > Replace with an explicit newline.
> >
> > Reported-by: Bruce Richardson
> > Signed-off-by: David Marchand
>
> Acked-by: Bruce Richardson
Applied
Thanks
--
Thomas
> > ENIC PMD was giving compilation errors on ppc_64-power8-linuxapp-gcc because
> > of types such as u_int32_t. This patch replaces all those with uint32_t and
> > similar ones.
> >
> > Reported-by: David Marchand
> > Signed-off-by: Sujith Sankar
>
> Acked-by: Bruce Richardson
>
> This
2014-10-23 16:36, Sergio Gonzalez Monroy:
> Building combined shared libraries results in applications being linked
> against separeted/individual and combined libs altogether.
>
> Link only against combined lib when the config option is enabled.
>
> Signed-off-by: Sergio Gonzalez Monroy
[...]
On Fri, Nov 28, 2014 at 10:38 AM, Sujith Sankar wrote:
> ENIC PMD was giving compilation errors on ppc_64-power8-linuxapp-gcc
> because
> of types such as u_int32_t. This patch replaces all those with uint32_t
> and
> similar ones.
>
> Reported-by: David Marchand
> Signed-off-by: Sujith Sankar
Hi Sergio,
2014-10-22 17:36, Sergio Gonzalez Monroy:
> Incompatible libraries error when building shared libraries for 32bits on
> a 64bits system.
> Fix issue by passing CPU_CFLAGS to CC when LINK_USING_CC is enabled.
This issue looks really strange. If that's the only way to fix it,
it would
This patch adds a new sample app UG, contaning explanation
of the new two sample apps added in the VM power management
patchset
Changes in v2:
Corrected svg files
Signed-off-by: Pablo de Lara
---
.../sample_app_ug/img/vm_power_mgr_highlevel.svg | 1173
Use of \n in sed expression is not portable and triggered an invalid
configuration on BSD (at least).
Replace with an explicit newline.
Reported-by: Bruce Richardson
Signed-off-by: David Marchand
---
scripts/gen-config-h.sh |6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff
On Fri, Nov 28, 2014 at 3:59 PM, Bruce Richardson <
bruce.richardson at intel.com> wrote:
> Yes, it's ugly, but it's probably more resilient. I'm looking forward to
> getting
> an option C.
>
Option C and D are ugly as well (using some bashism like $' ' or using an
intermediate variable with a
Hi Sergio, Neil,
Do you agree to add this comment (before applying this *needed* patch)?
+# Workaround lack of DT_NEEDED entry
> -EXECENV_LDFLAGS =
> +EXECENV_LDFLAGS = --no-as-needed
--
Thomas
> > The statistics that is reported through the rx_nombuf fields in struct
> > rte_eth_stats was not set when the vector PMD was used. The statistics
> > should report the number of mbufs that could _not_ be allocated during
> > rearm of the RX queue. The non-vector PMD reports it correctly. The
>
2014-11-28 15:10, Pablo de Lara:
> gcc 4.3 complains that slow_pkts array in bond_ethdev_tx_burst_8023ad may be
> used uninitialized, so it has been initialized to NULL
>
> Signed-off-by: Pablo de Lara
Acked-by: Thomas Monjalon
Applied
Thanks
--
Thomas
On 28/11/14 9:36 pm, "Bruce Richardson" wrote:
>On Fri, Nov 28, 2014 at 04:01:00PM +, Sujith Sankar (ssujith) wrote:
>>
>>
>> On 28/11/14 9:22 pm, "Bruce Richardson"
>>wrote:
>>
>> >On Fri, Nov 28, 2014 at 03:08:19PM +0530, Sujith Sankar wrote:
>> >> ENIC PMD was giving compilation
On 11/28/2014 01:01 AM, Thomas Monjalon wrote:
> 2014-10-31 15:37, Jincheng Miao:
>> Function pci_num_vf() is introduced from upstream linux-2.6.34. So
>> this patch make compatible with longterm kernel linux-2.6.32.63.
>>
>> For RHEL6's kernel, although it is based on linux-2.6.32, it has
>>
On Fri, Nov 28, 2014 at 04:01:00PM +, Sujith Sankar (ssujith) wrote:
>
>
> On 28/11/14 9:22 pm, "Bruce Richardson" wrote:
>
> >On Fri, Nov 28, 2014 at 03:08:19PM +0530, Sujith Sankar wrote:
> >> ENIC PMD was giving compilation errors on ppc_64-power8-linuxapp-gcc
> >>because
> >> of types
On 28/11/14 9:22 pm, "Bruce Richardson" wrote:
>On Fri, Nov 28, 2014 at 03:08:19PM +0530, Sujith Sankar wrote:
>> ENIC PMD was giving compilation errors on ppc_64-power8-linuxapp-gcc
>>because
>> of types such as u_int32_t. This patch replaces all those with
>>uint32_t and
>> similar ones.
>>
On Fri, Nov 28, 2014 at 03:08:19PM +0530, Sujith Sankar wrote:
> ENIC PMD was giving compilation errors on ppc_64-power8-linuxapp-gcc because
> of types such as u_int32_t. This patch replaces all those with uint32_t and
> similar ones.
>
> Reported-by: David Marchand
> Signed-off-by: Sujith
On Fri, Nov 28, 2014 at 04:42:44PM +0100, David Marchand wrote:
> Use of \n in sed expression is not portable and triggered an invalid
> configuration on BSD (at least).
> Replace with an explicit newline.
>
> Reported-by: Bruce Richardson
> Signed-off-by: David Marchand
Acked-by: Bruce
> -Original Message-
> From: Olivier MATZ [mailto:olivier.matz at 6wind.com]
> Sent: Friday, November 28, 2014 11:19 AM
> To: Ananyev, Konstantin; Liu, Jijiang; dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v3 1/4] mbuf:add three TX offload flags and
> change three fields
>
> Hi
Hello,
I have ~15Gbps of traffic flowing through two 10GE ports and been profiling
the rte mempool(or rather the pktmbuf mempool) memory consumption:
I have per lcore caching disabled(cache_size is 0)
I have noticed that:
- Mempool FREE cnt(as given byt rte_mempool_free_count()) increases
-
Hi Sujith,
Some tips when sending a v2:
- use --in-reply-to to keep all the versions in the same thread
- use --annotate to write a changelog
These guidelines are explained in this page:
http://dpdk.org/dev#send
Thanks
--
Thomas
2014-11-28 10:50, Olivier MATZ:
> Hi Jijiang,
>
> On 11/27/2014 06:03 PM, Jijiang Liu wrote:
> > @@ -303,7 +305,7 @@ process_outer_cksums(void *outer_l3_hdr, uint16_t
> > outer_ethertype,
> > * TESTPMD_TX_OFFLOAD_* in ports[tx_port].tx_ol_flags. They control
> > * wether a checksum must be
Hi Jijiang,
On 11/28/2014 09:54 AM, Liu, Jijiang wrote:
>>> My understanding of the meaning of the flags is:
>>>
>>>- PKT_TX_IP_CKSUM: tell the NIC to compute IP cksum
>
>> My initial thought:
>> It tells the NIC that it is an IPV4 packet for which it has to compute
>> checksum.
>>
>>>
>>>
Hi Jijiang,
On 11/27/2014 06:03 PM, Jijiang Liu wrote:
> @@ -303,7 +305,7 @@ process_outer_cksums(void *outer_l3_hdr, uint16_t
> outer_ethertype,
> * TESTPMD_TX_OFFLOAD_* in ports[tx_port].tx_ol_flags. They control
> * wether a checksum must be calculated in software or in hardware. The
>
On 11/27/2014 06:03 PM, Jijiang Liu wrote:
> It will avoid to send a packet with a bad info:
> - we receive a Ether/IP6/IP4/L4/data packet
> - the driver sets PKT_RX_IPV6_HDR
> - the stack decapsulates IP6
> - the stack sends the packet, it has the PKT_TX_IPV6 flag but it's an IPv4
>
Hello Sujith,
Why keep those u8, u16, u32 etc... ?
Especially, you can see in this patch that the ioread16 uses u16, while
ioread8, ioread32 uses uint*.
I like consistency and "standard" types (unless there is a reason why we
use different types).
Anyway, this patch builds fine on ppc.
--
Hi Helin,
On 11/28/2014 09:07 AM, Zhang, Helin wrote:
> After I have completed another task, I read the datasheet carefully again.
> For those 5
> error bits I introduced for a long time, I'd like to explain one by one as
> below.
>
> #define PKT_RX_EIP_CKSUM_BAD (0ULL << 0) /**< External IP
28.11.2014 3:04, Thomas Monjalon ?:
> 2014-11-20 11:15, Yerden Zhumabekov:
>> These patches bring a fallback mechanism to ensure that CRC32 hash is
>> calculated regardless of hardware support from CPU (i.e. SSE4.2 intrinsics).
>> Performance is also improved by slicing data in 8 bytes.
>>
Hi Olivier, Konstantin
> -Original Message-
> From: Olivier MATZ [mailto:olivier.matz at 6wind.com]
> Sent: Wednesday, November 26, 2014 10:12 PM
> To: Ananyev, Konstantin; Zhang, Helin; dev at dpdk.org
> Cc: Cao, Waterman; Cao, Min
> Subject: Re: [PATCH] i40e: Use one bit flag for all
On 28/11/14 1:54 am, "Thomas Monjalon" wrote:
>2014-11-27 19:01, Thomas Monjalon:
>> 2014-11-27 22:44, Sujith Sankar:
>> > Inclusion of vfio.h was giving compilation errors if kernel version
>>is less
>> > than 3.6.0 and if RTE_EAL_VFIO was on in config.
>> >
>> > Replaced inclusion of vfio.h
Hi all,
I have no comments on this issue, but I indeed see many places do have
this kernel issue(before/now/future), so can solve this issue globally?
Thus, we do not need to fix this case by case.
One solution(not sure if it works or not):
1. features and kernel version required list.
2. When
The changes include:
1. use the new introduced ol_flags and fields in csumonly.c file;
2. fix an issue of outer UDP checksum check;
3. change process logic in the process_inner_cksums();
Signed-off-by: Jijiang Liu
---
app/test-pmd/csumonly.c | 65
Rework the i40e PMD codes using the new introduced ol_flags and fields.
Signed-off-by: Jijiang Liu
---
lib/librte_pmd_i40e/i40e_rxtx.c | 52 +--
1 files changed, 28 insertions(+), 24 deletions(-)
diff --git a/lib/librte_pmd_i40e/i40e_rxtx.c
It will avoid to send a packet with a bad info:
- we receive a Ether/IP6/IP4/L4/data packet
- the driver sets PKT_RX_IPV6_HDR
- the stack decapsulates IP6
- the stack sends the packet, it has the PKT_TX_IPV6 flag but it's an IPv4
packet.
Signed-off-by: Jijiang Liu
---
In place of removing the PKT_TX_VXLAN_CKSUM, we introduce 3 new flags:
PKT_TX_OUTER_IP_CKSUM, PKT_TX_OUTER_IPV6 and PKT_TX_UDP_TUNNEL_PKT, and a new
field: l4_tun_len.
Replace the inner_l2_len and the inner_l3_len field with the outer_l2_len and
outer_l3_len field.
PKT_TX_OUTER_IP_CKSUM: is
We have got some feedback about backward compatibility of VXLAN TX checksum
offload API with 1G/10G NIC after the i40e VXLAN TX checksum codes were
applied, so we have to rework the APIs on i40e, including the changes of mbuf,
i40e PMD and csum engine.
The main changes in mbuf are as follows,
The changes include:
1. use the new introduced ol_flags and fields in csumonly.c file;
2. fix an issue of outer UDP checksum check;
3. change process logic in the process_inner_cksums();
Signed-off-by: Jijiang Liu
---
app/test-pmd/csumonly.c | 65
Rework the i40e PMD codes using the new introduced ol_flags and fields.
Signed-off-by: Jijiang Liu
---
lib/librte_pmd_i40e/i40e_rxtx.c | 49 +-
1 files changed, 27 insertions(+), 22 deletions(-)
diff --git a/lib/librte_pmd_i40e/i40e_rxtx.c
It will avoid to send a packet with a bad info:
- we receive a Ether/IP6/IP4/L4/data packet
- the driver sets PKT_RX_IPV6_HDR
- the stack decapsulates IP6
- the stack sends the packet, it has the PKT_TX_IPV6 flag but it's an IPv4
packet.
Signed-off-by: Jijiang Liu
---
In place of removing the PKT_TX_VXLAN_CKSUM, we introduce 3 new flags:
PKT_TX_OUTER_IP_CKSUM, PKT_TX_OUTER_IPV6 and PKT_TX_UDP_TUNNEL_PKT, and a new
field: l4_tun_len.
Replace the inner_l2_len and the inner_l3_len field with the outer_l2_len and
outer_l3_len field.
PKT_TX_OUTER_IP_CKSUM: is
We have got some feedback about backward compatibility of VXLAN TX checksum
offload API with 1G/10G NIC after the i40e VXLAN TX checksum codes were
applied, so we have to rework the APIs on i40e, including the changes of mbuf,
i40e PMD and csum engine.
The main changes in mbuf are as follows,
42 matches
Mail list logo
