[dpdk-dev] [PATCH] pci: don't insert an unbound device to pci_device_list in pci_scan_one

2016-06-25 Thread Rugang Chen
If a device isn't bound by any uio driver (vfio-pci, igb_uio, uio_pci_generic) and is expected to owned by a kernel space driver, here it's still inserted to pci_device_list. This may cause application based on dpdk fetch the device by accident and then the device is hanlded by dpdk. For safe, sk

[dpdk-dev] [PATCH] mempool: fix symbol export

2016-06-25 Thread Thomas Monjalon
Every new symbols in release 16.07 are exported with the version string DPDK_16.07. Also remove the empty local: section which is not needed because inherited from the DPDK_2.0 block. Signed-off-by: Thomas Monjalon --- lib/librte_mempool/rte_mempool_version.map | 3 +-- 1 file changed, 1 inserti

[dpdk-dev] [PATCH v2] cryptodev: uninline parameter parsing

2016-06-25 Thread Thomas Monjalon
There is no need to have this parsing inlined in the header. It brings kvargs dependency to every crypto drivers. The functions are moved into rte_cryptodev.c. Signed-off-by: Thomas Monjalon --- v2: - remove kvargs dep in null PMD - add function in .map file --- drivers/crypto/null/Makefile

[dpdk-dev] [PATCH] scripts: relax line length check for fixed commit

2016-06-25 Thread Wiles, Keith
On 6/24/16, 4:30 AM, "dev on behalf of Bruce Richardson" wrote: >On Fri, Jun 24, 2016 at 12:44:18AM +0200, Thomas Monjalon wrote: >> It is better to keep the line "Fixes:" longer than 75 characters >> than splitting. >> >> Signed-off-by: Thomas Monjalon > >Definite +1 Yes, +2 for me. > >Acked

[dpdk-dev] backtracing from within the code

2016-06-25 Thread Rosen, Rami
Hi, If you are willing to skip static methods and use the GCC backtrace, you can try this example (it worked for me, but it was quite a time ago): http://www.helicontech.co.il/?id=linuxbt Regards, Rami Rosen Intel Corporation -Original Message- From: dev [mailto:dev-boun...@dpdk.org] On

[dpdk-dev] [PATCH v2] cryptodev: uninline parameter parsing

2016-06-25 Thread De Lara Guarch, Pablo
Hi Thomas, > -Original Message- > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com] > Sent: Saturday, June 25, 2016 2:14 PM > To: Doherty, Declan; De Lara Guarch, Pablo > Cc: dev at dpdk.org > Subject: [PATCH v2] cryptodev: uninline parameter parsing > > There is no need to have

[dpdk-dev] [PATCH] mbuf:rearrange mbuf to be more mbuf chain friendly

2016-06-25 Thread Keith Wiles
Move the next pointer to the first cacheline of the rte_mbuf structure and move the offload values to the second cacheline to give better performance to applications using chained mbufs. Enabled by a configuration option CONFIG_RTE_MBUF_CHAIN_FRIENDLY default is set to No. Signed-off-by: Keith Wi

[dpdk-dev] [PATCH] ixgbe:enable configuration for old ptype behavior

2016-06-25 Thread Keith Wiles
The default behavior is to NOT support the old ptype behavior, but enabling the configuration option the old ptype style can be supported. Add support for old behaviour until we have a cleaner solution using a configuration option CONFIG_RTE_IXGBE_ENABLE_OLD_PTYPE_BEHAVIOUR, which is defaulted to

[dpdk-dev] [PATCH] mbuf:rearrange mbuf to be more mbuf chain friendly

2016-06-25 Thread Wiles, Keith
On 6/25/16, 10:29 AM, "dev on behalf of Keith Wiles" wrote: >Move the next pointer to the first cacheline of the rte_mbuf structure >and move the offload values to the second cacheline to give better >performance to applications using chained mbufs. > >Enabled by a configuration option CONFIG_RT

[dpdk-dev] [PATCH] librte_ether: Fix a typo in rte_ethdev.h

2016-06-25 Thread Rami Rosen
This patch fixes a typo in librte_ether/rte_ethdev.h. Signed-off-by: Rami Rosen --- lib/librte_ether/rte_ethdev.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/lib/librte_ether/rte_ethdev.h b/lib/librte_ether/rte_ethdev.h index 45482f1..2dab75c 100644 --- a/lib/librte_ether

[dpdk-dev] [PATCH v2] mbuf:rearrange mbuf to be more mbuf chain friendly

2016-06-25 Thread Keith Wiles
Move the next pointer to the first cacheline of the rte_mbuf structure and move the offload values to the second cacheline to give better performance to applications using chained mbufs. Enabled by a configuration option CONFIG_RTE_MBUF_CHAIN_FRIENDLY default is set to No. Signed-off-by: Keith Wi

[dpdk-dev] [PATCH] app/test: avoid freeing mbuf twice

2016-06-25 Thread Pablo de Lara
In cryptodev tests, when input and output buffers were the same, the mbuf was being freed twice, causing refcnt_atomic to be negative. Fixes: 202d375c60bc ("app/test: add cryptodev unit and performance tests") Signed-off-by: Pablo de Lara --- app/test/test_cryptodev.c | 5 - 1 file changed,

[dpdk-dev] [PATCH] mempool: fix symbol export

2016-06-25 Thread Olivier Matz
On 06/25/2016 03:10 PM, Thomas Monjalon wrote: > Every new symbols in release 16.07 are exported with the version > string DPDK_16.07. > Also remove the empty local: section which is not needed because > inherited from the DPDK_2.0 block. > > Signed-off-by: Thomas Monjalon > --- > lib/librte_m

[dpdk-dev] [PATCH] pci:don't insert an unbound device to pci_device_list in pci_scan_one

2016-06-25 Thread Rugang Chen
Hi all, Can you take a look at the patch and send out your comments? It's first time for me to work with git as this way, not sure if I'm doing right on the process. There're two emails for this patch. The second one is to correct the first one. On Saturday, June 25, 2016, Rugang Chen wrote: