[lng-odp] Regarding ODP rings in linux-generic

Hi, I have a usecase where each core has a ring associated with it. Every core dequeues from its own ring - thus each core's ring above is representing the single consumer mode. Every core can enqueue to every other core's ring - thus each core's ring above is representing the multi-producer mod

Re: [lng-odp] Regarding ODP rings in linux-generic

rings in linux-generic Hi Prashant, do you pass flags to odp_ring_create that it's multi consumer, multi producer? Maxim. On 08/22/2014 12:12 PM, Prashant Upadhyaya wrote: > > Hi, > > I have a usecase where each core has a ring associated with it. > > Every core deque

[lng-odp] ODP Compilation for Cavium Octeon 6645

Hi, I was trying to compile ODP (v0.2) for octeon and ran into the following compilation issue - /tmp/ccH0X8h8.s: Assembler messages: /tmp/ccH0X8h8.s:44: Error: opcode not supported on this processor: octeon (mips64r2) `lai $2,($4)' It seems the error comes from the file odp_atomic.h which has

Re: [lng-odp] ODP Compilation for Cavium Octeon 6645

Hi Jacob, I am using a version of SDK2.3 supplied by Adax. . ./env-setup OCTEON_CN66XX_PASS1_2 Compiler version is 4.3 mips64-octeon-linux-gnu-gcc Regards -Prashant From: Jacob, Jerin [mailto:jerin.ja...@caviumnetworks.com] Sent: Tuesday, September 16, 2014 2:43 PM To: Prashant Upadhyaya Cc

Re: [lng-odp] ODP Compilation for Cavium Octeon 6645

Hi Bala, Thanks, the addition of -march=octeon2 did the trick. So are we saying that we cannot compile for octeon1 ? Regards -Prashant From: Bala Manoharan [mailto:bala.manoha...@linaro.org] Sent: Wednesday, September 17, 2014 8:51 AM To: Prashant Upadhyaya Cc: Jacob, Jerin; lng-odp

Re: [lng-odp] odp_pktio_

Hi guys, Why would you want to deprecate odp_pktio_ This maps so neatly over the 'basic' way the packets can be obtained from hardware and sent to the hardware. What if I don't want to use queues and just want an interface to get packets into my application and send from there. The odp_pktio_ is

[lng-odp] Regarding 'atomic' functions

Hi, I am migrating my application code from ODP v0.2 to v0.7 In v0.2, there used to be a function like the following - odp_atomic_cmpset_u32 How do I realize the above in v0.7 ? Regards -Prashant "DISCLAIMER: This message is proprietary to Aricent and is intended solely for the use of the indi

[lng-odp] Question on odp_packet_t in linux generic v0.7

Hi, If I allocate a packet using odp_packet_alloc, and obtain a pointer to the data payload of the packet using odp_packet_data (assuming that l2_offset is set to zero), then is this packet pointer guaranteed to be cache aligned or not. Can somebody point me to the design spec of linux generic

[lng-odp] Question on frame_len of odp_packet_t

Hi, I am allocating a packet with odp_packet_alloc. Now I can write some data into it using odp_packet_copydata_in. Or even by getting the buffer payload first by odp_packet_data call. But now how do I change the frame_len of this odp_packet_t ? I see a function static inline void packet_set_len

[lng-odp] Regarding odp_packet_t and buffer chaining

Hi, Suppose I have pkt1 and pk2 which are instances of odp_packet_t. Now I have two questions - 1. what is the way to chain the buffers of pkt1 and pkt2 into pkt1 so that from now on I can just use pkt1 for transmission via, say, odp_pktio_send API 2. if I create a new odp_packet_

[lng-odp] Question on ODP-Netmap repository

Hi, In v0.2 the packet socket and netmap backends were part of the main repository. Now I see that there is a 'main' repository whose linux generic backend has packet sockets. Netmap has gone into a different repository. Now my question is - is the ODP-Netmap a standalone repo i.e when I build i

[lng-odp] Regarding ODP_SHM_NUM_BLOCKS

Hi, The file odp_shared_memory.c contains the following - #define ODP_SHM_NUM_BLOCKS 32 I have an application which creates one ring per core. Recently I was porting my application on a 48 core processor and the creation fails due to this limit on the number of blocks. A value of 64 works well

[lng-odp] [PATCH] Fix in linux-generic to increase value of ODP_SHM_NUM_BLOCKS

This fix changes the value of ODP_SHM_NUM_BLOCKS from 32 to 64 which might be useful for applications operating with high number of cores and creating per core blocks. Signed-off-by: Prashant Upadhyaya mailto:prashant.upadhy...@aricent.com>> -- diff -rupN v0.9.0/platform/linux-g

[lng-odp] [PATCH] Fix in linux-generic to increase value of ODP_SHM_NUM_BLOCKS

This fix changes the value of ODP_SHM_NUM_BLOCKS from 32 to 64 which might be useful for applications operating with high number of cores and creating per core blocks. Signed-off-by: Prashant Upadhyaya mailto:prashant.upadhy...@aricent.com>> --- platform/linux-generic/odp_shared_memory

[lng-odp] Regarding ODP Buffers

Hi, Suppose I create a buffer pool with odp_buffer_pool_create. Now I can allocate buffers from this pool with odp_packet_alloc. I can get hold of the packet data pointer with odp_packet_data. So far so good. Now, the buffer address which I obtain with odp_packet_data is a 'virtual' address. Fir

Re: [lng-odp] [PATCHv2 1/3] linux-generic: pktio: remove basic socket implementation

Hi guys, I object too. Why would you want to 'remove' a functionality. The 'basic' socket is a very good lowest common denominator of implementation across all the machines/linux And it was anyway a fallback mechanism. Currently atleast my application depends on it, so guys, please don't remove