2015-07-29 11:54, Yong Liu:
> From: Marvin Liu
>
> '--no-huge' option now can workable with -m option.
> Unit test for eal flag should change pass criterion.
>
> Signed-off-by: Marvin Liu
Fixes: a7de7e6beb69 ("eal: allow combining -m and --no-huge")
Applied, thanks
> > From: Marvin Liu
> >
> > In previous setting, mempool size and cache_size are both 32.
> > This is not satisfied with cache_size checking rule by now.
> > Cache size should less than CONFIG_RTE_MEMPOOL_CACHE_MAX_SIZE and
> > mempool size / 1.5.
> >
> > Signed-off-by: Marvin Liu
> >
> Acked
2015-08-03 17:28, Ananyev, Konstantin:
> > > 2015-06-26 10:07, Tomasz Kulasek:
> > > > This patch adds management of PKT_RX_FDIR and PKT_RX_RSS_HASH ol_flags
> > > > in vPMD
> > > > for unified packet type as well as for 16 bit field packet_type when
> > > > RTE_UNIFIED_PKT_TYPE is not defined.
>
2015-08-03 18:59, Konstantin Ananyev:
> Niantic HW expects Header Buffer Address in the RXD to be word aligned.
> So, if mbuf's buf_physaddr is not word aligned then
> RX path will not work properly.
> Right now, in ixgbe PMD we always setup Packet Buffer Address(PBA) and
> Header Buffer Address (H
On 2015/8/3 17:06, Thomas Monjalon wrote:
> 2015-08-03 14:31, Chao Zhu:
>> The rte_cpu_to_le_32 function can't be used to define const variables
>> because it has different implementation on big endian platforms. If
>> doing so, it will cause 'initializer element is not constant' compiling
>> err
2015-08-02 22:04, Matthew Hall:
> On Mon, Aug 03, 2015 at 09:46:54AM +0800, Liang, Cunming wrote:
> > According to the API definition, if the socket could not be determined, a
> > default of zero will take.
> > The '-1' is returned when the port_id value is out of range.
>
> Yes, but when I asked
2015-08-03 17:02, Zhang, Helin:
> I don't think so, as l3fwd cannot support all cases quite well at this moment.
> Can we targeting to do that in R2.2?
OK, thanks
> > -Original Message-
> > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> > Sent: Monday, August 3, 2015 9:55 A
Niantic HW expects Header Buffer Address in the RXD to be word aligned.
So, if mbuf's buf_physaddr is not word aligned then
RX path will not work properly.
Right now, in ixgbe PMD we always setup Packet Buffer Address(PBA) and
Header Buffer Address (HBA) to the same value:
buf_physaddr + RTE_PKTMBU
Any news?
Can we remove it from version 2.1?
2015-07-22 14:51, Zhang, Helin:
> Marvin/Waterman
>
> Could you help to check if l3fwd is good enough for all cases (1g/10/40g, PF
> and VF, single queue/multiple queue)?
> We aim to remove l3fwd-vf to reduce an example application which is not so
>
Helin, your review is needed.
Please check how to write a simple title.
2015-07-09 14:09, Zhe Tao:
> Issue: When calling the reset API to reset all the stats for i40e,
> all the stats will be untracked before the next time calling get stats API.
> Once calling the get stats API it will show all th
> > The function documentation was obviously copied and not updated.
> >
> > Signed-off-by: Stephen Hemminger
> Acked-by: Bruce Richardson
Applied, thanks
Patch 1/2 has been rejected.
2015-08-03 18:12, Nikita Kozlov:
> This option permit to build librte_kni.so without building rte_kni.ko so
> you can build a sdk without building kernel drivers.
>
> Signed-off-by: Nikita Kozlov
Applied, thannks
2015-08-03 16:16, Montorsi, Francesco:
> Hi all,
> I have searched the archives for this, without much success.
>
> Is it possible to build dpdk user-space libraries with -O0 and -g instead of
> -O3 ?
> This would make debugging via GDB much more friendly...
The answer is EXTRA_CFLAGS :)
This option permit to build librte_kni.so without building rte_kni.ko so
you can build a sdk without building kernel drivers.
Signed-off-by: Nikita Kozlov
---
config/common_linuxapp | 1 +
lib/librte_eal/linuxapp/Makefile | 2 +-
2 files changed, 2 insertions(+), 1 deletion(-)
diff --
2015-04-13 15:11, Stephen Hemminger:
> Part of my patch-a-day seeries
>
> Stephen Hemminger (2):
> pci: cleanup whitespace
> pci: rearrange logic from compare loop
>
> lib/librte_eal/linuxapp/eal/eal_pci.c | 41
> +--
> 1 file changed, 20 insertions(+), 21 de
2015-07-30 15:41, Thomas Monjalon:
> Build log:
> lib/librte_eal/bsdapp/eal/eal_pci.c:462:9: error:
> incompatible integer to pointer conversion passing 'u_int32_t'
> (aka 'unsigned int') to parameter of type 'void *'
>
> It is fixed by passing the pointer of pi.pi_data to memcpy.
>
> By the way,
Stephen,
> By the way, it seems strange that pi_data is initialized twice:
> .pi_data = *(u_int32_t *)buf
> memcpy(&pi.pi_data, buf, len);
Please, could you explain why there are 2 initializations of pi_data
in lib/librte_eal/bsdapp/eal/eal_pci.c?
2015-07-28 17:47, Thomas Monjalon:
> The basic build tests fail in many ways.
> The driver was disabled by error in first test before applying it.
> That's why it has been accepted.
>
> This is a first try to fix most of the errors.
> It has not been tested with old compilers, nor icc.
> The 32-bi
2015-07-28 17:47, Thomas Monjalon:
> Example of errors:
> error: cast to pointer from integer of different size
> [-Werror=int-to-pointer-cast]
> error: format ?%lx? expects argument of type ?long unsigned int?, but
> argument 5 has type ?uint64_t
> error: format ?%lu? expects a
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Zhe Tao
> Sent: Wednesday, July 8, 2015 11:09 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH] i40e: Fix the issue when resetting the stats for
> i40e,
> the stats information will be lost before next ti
> -Original Message-
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Monday, August 03, 2015 4:12 PM
> To: Ananyev, Konstantin
> Cc: dev at dpdk.org; Kulasek, TomaszX
> Subject: Re: [dpdk-dev] [PATCH v2] vpmd: rss hash ol_flag for unified packet
> type
>
> Ping
>
Ping
2015-07-10 14:24, Thomas Monjalon:
> Alan, Pablo, any comment?
> By the way, we need a maintainer for rte_power.
>
> 2015-07-07 16:36, Yong Liu:
> > From: Marvin Liu
> >
> > Two socket Haswell cpus on cottonwoodPass have 72 cores totally. And this
> > number oversize the default value of m
Hi, wenzhuo,
Have to check that because I really had the panic when compile my code
with dpdk 2.0.0.
Also I checked the code in the dpdk git repo. It does not check the
reference count, it just only calculate
the begin positions of mbufs that really needs to be recycled.
Ping
2015-07-16 17:40, Thomas Monjalon:
> Konstantin, what is the status of this patch?
>
> 2015-06-26 10:07, Tomasz Kulasek:
> > This patch adds management of PKT_RX_FDIR and PKT_RX_RSS_HASH ol_flags in
> > vPMD
> > for unified packet type as well as for 16 bit field packet_type when
> > RTE_UN
2015-06-29 11:28, Ananyev, Konstantin:
> Hi Eric,
>
> > -Original Message-
> > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Eric Kinzie
> > Sent: Thursday, June 25, 2015 7:26 PM
> > To: dev at dpdk.org
> > Subject: [dpdk-dev] [PATCH 2/2] ixgbe: add memory barriers in vector rx/t
2015-06-25 11:25, Eric Kinzie:
> zero values in ixgbe_reset_rx_queue() used by vector receive so that
> rearming the rx queue happens at the right time. Not doing so can in
> some cases result in the software inadvertently setting the card's rx
> tail pointer equal to the head pointer, which indic
Got it. Thanks!
On 2015/8/3 16:53, Thomas Monjalon wrote:
> 2015-08-03 14:31, Chao Zhu:
>> The using of rte_cpu_to_le_32() in pre-compile macros will cause error
>> 'initializer element is not constant' on big endian platforms. This patch
>> fixes the compilation error of fm10k driver.
>>
>>
>> Ch
I don't think so, as l3fwd cannot support all cases quite well at this moment.
Can we targeting to do that in R2.2?
Regards,
Helin
> -Original Message-
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Monday, August 3, 2015 9:55 AM
> To: Zhang, Helin
> Cc: Liu, Yong;
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen Hemminger
> Sent: Friday, July 24, 2015 10:06 AM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH 01/11] eth: move queue check for rx_queue_count
>
> All implemenations of rx_queue_count need to chec
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen Hemminger
> Sent: Friday, July 24, 2015 10:06 AM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH 10/11] i40e: remove useless cast
>
> drivers/net/i40e/i40e_ethdev.c:2360:18-56: WARNING: casting valu
> > In Rx and Tx queue_disable functions, the index of queue should
> > be qnum other than i which is the iteration of time expiration.
> > When a Tx queue fails to start in fm10k_dev_start, all Rx queues
> > and Tx queues that are started should be cleaned before the
> > function returns an error.
Oh, sorry, my mistakes, it is in the tx_release_mbuf
static void
ixgbe_tx_queue_release_mbufs(struct ixgbe_tx_queue *txq)
{
unsigned i;
if (txq->sw_ring != NULL) {
for (i = 0; i < txq->nb_tx_desc; i++) {
if (txq->sw_ring[i].mbuf != NULL) {
Hi all,
I have searched the archives for this, without much success.
Is it possible to build dpdk user-space libraries with -O0 and -g instead of
-O3 ?
This would make debugging via GDB much more friendly...
Thanks!
Francesco Montorsi
On IBM POWER8 PPC64 little endian architecture, the definition of tsc
union will be different. This patch fix this to enable the right output
from rte_rdtsc().
Signed-off-by: Chao Zhu
---
.../common/include/arch/ppc_64/rte_cycles.h|5 +
1 files changed, 5 insertions(+), 0 deletio
This patch add the definiton of tsc union for POWER8 PPC64 little endian
architecture.
Chao Zhu (1):
PPC64: add cpu cycle support to IBM POWER8 PPC64LE
.../common/include/arch/ppc_64/rte_cycles.h|5 +
1 files changed, 5 insertions(+), 0 deletions(-)
Hi Wenzhuo,
The issue is in the function *ixgbe_dev_free_queues* called in the
*ixgbe_dev_close*.
The *ixgbe_dev_free_queues* will call *ixgbe_rx_queue_release_mbuf* to
recycle all the mbuf on
the queues. If some mbufs have already been recycled by the
*ixgbe_tx_free_bufs*, the
> > Fixes issue where ieee15888 timestamping doesn't work for the i40e pmd when
> > RTE_ABI_NEXT is enabled.
> >
> > Also refactors repeated ieee15888 flag checking and setting code into a
> > function.
> >
> > Reported-by: Huilong Xu
> > Signed-off-by: John McNamara
> Acked-by: Wenzhuo Lu
A
> > Ieee1588 reads system time to set its timestamp. On 1G NICs, for example,
> > i350, system time is disabled by default. It means the ieee1588 timestamp
> > will always be 0.
> > This patch enables system time when ieee1588 is enabled.
> >
> > v2 changes:
> > * move the macro to the top of the
The rte_cpu_to_le_32 function can't be used to define const variables
because it has different implementation on big endian platforms. If
doing so, it will cause 'initializer element is not constant' compiling
error. This patch fixes this problem.
Signed-off-by: Chao Zhu
---
drivers/net/fm10k/ba
The using of rte_cpu_to_le_32() in pre-compile macros will cause error
'initializer element is not constant' on big endian platforms. This patch
fixes the compilation error of fm10k driver.
Chao Zhu (1):
fm10k: fix the compilation on big endian platforms
drivers/net/fm10k/base/fm10k_tlv.c |
Hello:
How can I configure my NIC setting(s) when loaded with the dpdk:igb_uio driver?
E.g I need to keep the MAC address and the IP address of the NIC the "same as
before" after switching to a dpdk:igb_uio driver for the NIC.
Is there a equiv of "ifconfig" utility in dpdk?
How do I addres
> > When adding a "depth small" entry, if its extended flag is not set and
> > its depth is smaller than the one in the tbl24, nothing should be done
> > otherwise will operate on the wrong memory area.
> >
> > Signed-off-by: Zhe Tao
>
> Acked-by: Cunming Liang
Applied, thanks
2015-08-03 19:32, Chao Zhu:
> On 2015/8/3 17:06, Thomas Monjalon wrote:
> > 2015-08-03 14:31, Chao Zhu:
> >> The rte_cpu_to_le_32 function can't be used to define const variables
> >> because it has different implementation on big endian platforms. If
> >> doing so, it will cause 'initializer eleme
2015-08-03 15:16, Chao Zhu:
> On IBM POWER8 PPC64 little endian architecture, the definition of tsc
> union will be different. This patch fix this to enable the right output
> from rte_rdtsc().
>
> Signed-off-by: Chao Zhu
Applied, thanks
> Fix config options for ixgbe, i40e and KNI for BSD.
>
> Thomas Monjalon (3):
> ixgbe: fix offload config option name
> config: enable same drivers options for linux and bsd
> config: remove kni options for bsd
Applied
ping
On 24.07.2015 15:59, Zoltan Kiss wrote:
> Hi,
>
> On 24/07/15 11:13, Ciprian Barbu wrote:
>> From: Ciprian Barbu
>>
>> This tries to show an approximate behavior described in an earlier
>> discussion
>> called "can eth_igb_xmit_pkts called with len 0 affect transmission?"
>>
>> I'm using Int
I Have Implemented the ACL Rules from your sample dpdk application in my
Application.
Though I am Not gettig any Error/Warning But the rte_acl_classify() not
returning any response to the results .
Kindly help me.
Confirmed. It can compile on Power8 Big Endian.
Thank you!
On 2015/8/3 10:39, Chao Zhu wrote:
>
> Really sorry for the delay.
> Originally, I thought the email was to asking the ABI checking tools
> on Power which I'm not so familiar with. So this took me some time to
> find solution. For Power
2015-08-03 14:31, Chao Zhu:
> The rte_cpu_to_le_32 function can't be used to define const variables
> because it has different implementation on big endian platforms. If
> doing so, it will cause 'initializer element is not constant' compiling
> error. This patch fixes this problem.
> --- a/driver
2015-08-03 14:31, Chao Zhu:
> The using of rte_cpu_to_le_32() in pre-compile macros will cause error
> 'initializer element is not constant' on big endian platforms. This patch
> fixes the compilation error of fm10k driver.
>
>
> Chao Zhu (1):
> fm10k: fix the compilation on big endian platform
Chao,
The original need was to check with the ABI checker tool that the ABI
was not broken on POWER BE by this patch.
2015-08-03 11:45, Chao Zhu:
> Confirmed. It can compile on Power8 Big Endian.
> Thank you!
>
> On 2015/8/3 10:39, Chao Zhu wrote:
> >
> > Really sorry for the delay.
> > Original
Really sorry for the delay.
Originally, I thought the email was to asking the ABI checking tools on
Power which I'm not so familiar with. So this took me some time to find
solution. For Power little endian, the build is OK. I'll give feedback
when I tried Big endian compilation.
On 2015/7/31
Hi Peng,
As Wenzhuo pointed, we have a special version of tx_release_mbufs fo vector TX:
ixgbe_tx_queue_release_mbufs_vec().
It only frees sw_ring[] entries that contains valid mbufs.
So we don't need to set sw_ring[].mbuf = NULL at ixgbe_tx_free_bufs().
About the panic you mentioned: I think tha
Hi,
[...]
> > Another thing, that I just thought about:
> > Right now we invoke ixgbe_rxq_rearm() only at the start of
> > _recv_raw_pkts_vec().
> > Before it was ok, as _recv_raw_pkts_vec() would never try to read more then
> 32
> > RXDs.
> > But what would happen if nb_pkts > rxq->nb_desc and rx
Hi,
On 8/1/2015 11:56 AM, Matthew Hall wrote:
> I asked about this many months ago and was informed that "-1" is a "standard
> error value" that I should expect from these APIs when NUMA is not present.
> Now we're saying I have to change my code again to handle a zero value?
>
> Also not sure how
On Mon, Aug 03, 2015 at 10:41:47AM +0200, Thomas Monjalon wrote:
> Chao,
> The original need was to check with the ABI checker tool that the ABI
> was not broken on POWER BE by this patch.
>
Yes, its not compilation thats the concern, its ABI compatibility.
Neil
>
> 2015-08-03 11:45, Chao Zhu:
Hi Peng,
There're 2 versions of ixgbe_tx_queue_release_mbufs. One in ixgbe_rxtx.c, the
other is in ixgbe_rxtx_vec.c.
What you shown is the one in ixgbe_rxtx.c. You can find the one in
ixgbe_rxtx_vec.c already has the sanity check for reference count.
Thanks.
Hi Peng,
> -Original Message-
> From: HePeng [mailto:xnhp0320 at icloud.com]
> Sent: Monday, August 3, 2015 3:09 PM
> To: Lu, Wenzhuo
> Cc: dev at dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] [new]ixgbe:set txep.mbuf to NULL when calling
> ixgbe_tx_free_bufs
>
> Hi Wenzhuo,
> The issu
Hi,
[...]
> > > uint16_t
> > > ixgbe_recv_scattered_pkts_vec(void *rx_queue, struct rte_mbuf **rx_pkts,
> > > uint16_t nb_pkts)
> > > {
> > > struct ixgbe_rx_queue *rxq = rx_queue;
> > > - uint8_t split_flags[RTE_IXGBE_VPMD_RX_BURST] = {0};
> > > + uint8_t split_
Tested-by:huilong xu
Os: dpdk-fedora20 3.11.10-301.fc20.x86_64
Gcc: 4.8.3 20140911 (Red Hat 4.8.3-7)
NIC: fortville_spirit_single
Package:dpdk.org newest code(commit 921353134fe01606411d3508f3ca798f0a38da73) +
patch
Test case :ieee 1588 disable and ieee 1588 enable all passed.
Test steps:
1. enab
Hi Peng,
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of hepeng
> Sent: Saturday, August 1, 2015 9:27 AM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH] [new]ixgbe:set txep.mbuf to NULL when calling
> ixgbe_tx_free_bufs
>
> In *ixgbe_tx_free_bufs*, after
Hi,
[...]
> > diff --git a/drivers/net/ixgbe/ixgbe_rxtx.c b/drivers/net/ixgbe/ixgbe_rxtx.c
> > index 3f808b3..dbdb761 100644
> > --- a/drivers/net/ixgbe/ixgbe_rxtx.c
> > +++ b/drivers/net/ixgbe/ixgbe_rxtx.c
> > @@ -4008,7 +4008,8 @@ ixgbe_set_rx_function(struct rte_eth_dev *dev)
> > */
> >
Thanks Thomas.
> -Original Message-
> From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com]
> Sent: Monday, August 03, 2015 5:27 AM
> To: Liu, Yong
> Cc: dev at dpdk.org; olivier.matz at 6wind.com; Wu, Jingjing
> Subject: Re: [dpdk-dev] [PATCH] app test: fix mempool cache_size not ma
Hi,
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of John McNamara
> Sent: Friday, July 31, 2015 7:39 PM
> To: dev at dpdk.org
> Subject: [dpdk-dev] [PATCH v2] i40e: fix for ieee15888 with rte_next_abi
>
> Fixes issue where ieee15888 timestamping doesn't work
> > It looks better to have a data buffer address that is aligned to
> > 8 bytes. This is the case when there is no mbuf private area, but
> > if there is one, the alignment depends on the size of this area
> > that is located between the mbuf structure and the data buffer.
> >
> > Indeed, some dr
2015-07-29 18:12, Ananyev, Konstantin:
> As we don't support split header feature anyway, I think we can fix it just
> by always setting HBA in the RXD to zero.
> Could you try the fix for ixgbe below?
Please Konstantin, could you send a proper patch for this fix?
Thanks
> --- a/drivers/net/ixg
2015-07-30 22:26, Olivier Matz:
> Compiling for dpdk x86_x32 gives the following error:
>
> CC eal_common_timer.o
> In file included from /usr/include/sys/sysctl.h:63:0,
> from dpdk.org/lib/librte_eal/common/eal_common_timer.c:39:
> /usr/include/bits/sysctl.h:19:3: error: #error
2015-07-31 15:14, Nelio Laranjeiro:
> librte_pmd_mlx4.so needs to be linked with libibverbs otherwise, the PMD is
> not able to open Mellanox devices and the following message is printed by
> testpmd at startup
> "librte_pmd_mlx4: cannot access device, is mlx4_ib loaded?".
>
> Applications depende
2015-07-27 18:46, rsanford2 at gmail.com:
> From: Robert Sanford
>
> This patchset fixes a bug in timer stress test 2, adds a new stress test
> to expose a race condition bug in API rte_timer_manage(), and then fixes
> the rte_timer_manage() bug.
>
> Description of rte_timer_manage() race condit
69 matches
Mail list logo