In order to support more than 256 virtual ports, the field "port" in
rte_mbuf has been increased to 16 bits. The initialization/reset
value of the field "port" should be changed from 0xff to 0x
accordingly.
Signed-off-by: Zhiyong Yang
---
lib/librte_mbuf/rte_mbuf.c | 2 +-
lib/librte_mbuf/rt
-Original Message-
> Date: Fri, 11 Aug 2017 10:55:38 +0530
> From: "Rao, Nikhil"
> To: Jerin Jacob
> CC: gage.e...@intel.com, dev@dpdk.org, tho...@monjalon.net,
> bruce.richard...@intel.com, harry.van.haa...@intel.com,
> hemant.agra...@nxp.com, nipun.gu...@nxp.com, narender.vang...@inte
Yes , off course .
In my next patch the rdma-core is master
f11292efd54132e09c9586a72c6c7bd0f341659d.
I will add that to the patch in the cover letter.
-Shachar Beiser.
shacharbe@pegasus08-006:~/rdma-core$ git remote --v
origin http://l-gerrit.mtl.labs.mlnx:8080/upstream/rdma-core.git
-Original Message-
> Date: Wed, 30 Aug 2017 14:31:07 +0100
> From: Ferruh Yigit
> To: Jerin Jacob , dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH] net/thunderx: add device speed capability
> info
> User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101
> Thunderbird/52.
ping???
From: Thomas Monjalon
Sent: Tuesday, August 1, 2017 6:12:26 PM
To: Dumitrescu, Cristian
Cc: dev@dpdk.org; Prathyusha, Guduri
Subject: Re: [dpdk-dev] [PATCH] librte_cfgfile: add section num entries by index
01/08/2017 14:36, Dumitrescu, Cristian:
> > 27/04
-Original Message-
> Date: Wed, 9 Aug 2017 14:58:04 -0500
> From: Gage Eads
> To: dev@dpdk.org
> CC: jerin.ja...@caviumnetworks.com, nipun.gu...@nxp.com,
> hemant.agra...@nxp.com, harry.van.haa...@intel.com,
> bruce.richard...@intel.com
> Subject: [PATCH] eventdev: ease single-link queue
The reasons to modify RTE_MAX_ETHPORTS is the following.
1. RTE_MAX_ETHPORTS=32 by default has not met user's requirements
with development of virtualization technology. Some vdev users have
to modify the setting before the compiling.
2. port_id have been extended to 16 bits definition. But for m
The currnet mechanisim of allocating mbuf depends on RTE_MAX_ETHPORTS
which is hardcoded at the compiling time. Once large numbers of ports
are needed by users, it may easily cause the failure of creating
mempool due to be lack of enough hugepage memory.
The patch introduces the policy to limit th
Modify port_id related code in examples accordingly since port_id
definition in lib and pmd changes.
Fix some original checkpatch issues and remove some unnecessary
cast at the same time.
Signed-off-by: Zhiyong Yang
---
doc/guides/rel_notes/release_17_11.rst | 4 +
examples/bond/m
port_id is currently defined as uint8_t, which is limited to the
range 0 to 255. A larger range is required for vdev scalability.
It is necessary for a redefinition of port_id to extend it from
1 bytes to 2 bytes. All ethdev APIs and usages related to port_id
will be changed at the same time.
Dis
-Original Message-
> Date: Mon, 24 Jul 2017 13:48:20 +0100
> From: Harry van Haaren
> To: dev@dpdk.org
> CC: jerin.ja...@caviumnetworks.com, Harry van Haaren
>
> Subject: [PATCH] eventdev: add dev id checks to config functions
> X-Mailer: git-send-email 2.7.4
>
> This commit adds checks
Hi Konstantin,
About the IP identifier, I check the linux codes and have some feedbacks inline.
On Wed, Aug 30, 2017 at 09:38:33AM +0800, Ananyev, Konstantin wrote:
>
>
> > -Original Message-
> > From: Hu, Jiayu
> > Sent: Thursday, August 24, 2017 3:16 PM
> > To: dev@dpdk.org
> > Cc: Ka
> -Original Message-
> From: Wu, Jingjing
> Sent: Thursday, August 24, 2017 10:11 AM
> To: Tan, Jianfeng
> Cc: Wu, Jingjing; dev@dpdk.org
> Subject: [PATCH v3] eal: add counter size for efd clean
>
> For virtual device, the rte_intr_handle struct is
> initialized by the virtual device dr
I've skimmed the bug. It seems that it has problems to call the implementation
codes and only call the resolver codes.
But in fact, I only use FMV to avoid compilation issues (example: AVX512 aren't
support if doesn't have compilation option -mavx512).
I just add the attribution such as " __attri
Signed-off-by: Jingjing Wu
---
drivers/net/i40e/base/i40e_osdep.h | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/net/i40e/base/i40e_osdep.h
b/drivers/net/i40e/base/i40e_osdep.h
index c57ecde..fbcb729 100644
--- a/drivers/net/i40e/base/i40e_osdep.h
+++ b/drivers/net/i40e/base/i40e_osd
I didn't consider that the orginal codes are inline, would be replaced in
compilation time and have no function calls.
Anyway, I'll do perf test this week. And if it drops a lot, I will change to
codes like rte_memcpy_AVX512 (less function calls) and test perf again.
Best Regards,
Xiaoyun Li
On Sun, 3 Sep 2017 23:49:44 +0800
Jeff Guo wrote:
> + /**
> + * add device uevent file descriptor
> + * into wait list for uevent monitoring.
> + */
> + ev.events = EPOLLIN | EPOLLPRI | EPOLLRD
On Sun, 3 Sep 2017 23:49:44 +0800
Jeff Guo wrote:
> + char buf[RTE_EAL_UEVENT_MSG_LEN];
> +
> + memset(uevent, 0, sizeof(struct rte_eal_uevent));
> + memset(buf, 0, RTE_EAL_UEVENT_MSG_LEN);
Please don't initialize everything all the time; you are even initializing
receive data.
>
On Sun, 3 Sep 2017 23:49:44 +0800
Jeff Guo wrote:
> +int
> +rte_eal_uev_enable(int netlink_fd)
> +{
> + struct sockaddr_nl addr;
> + int ret;
> + int size = 64 * 1024;
> + int nonblock = 1;
> + memset(&addr, 0, sizeof(addr));
Blank line between declarations and code.
Also us
On Sun, 3 Sep 2017 23:49:44 +0800
Jeff Guo wrote:
> +int
> +rte_eal_uev_fd_new(void)
> +{
> +
> + int netlink_fd = -1;
> +
> + netlink_fd = socket(PF
Please don't use the "initialize everything" style of programming.
Gcc has good detection and warning about uninitalized variables, and
use testpmd for example, to show app how to request and use
uevent monitoring to handle an event of device hot removal.
Signed-off-by: Jeff Guo
---
app/test-pmd/testpmd.c | 62 ++
1 file changed, 62 insertions(+)
diff --git a/app/test-pmd/testpmd.
So far, about hot plug in dpdk, we already have hot plug add/remove
api and fail-safe driver to offload the fail-safe work from the app
user. But there are still lack of a general event api, since the interrupt
event, which hot plug related with, is diversity between each device and
driver, such as
This patch aim to add a general uevent mechanism in eal device layer,
to enable all kernel object hot plug monitoring, so user could use these
API to monitor and read out the device status info sent from the kernel
side, then corresponding to handle it, such as detach or attach the
device, and even
The corrupted code don't reply error in case of MAC
address adding failure while failsafe PMD was trying
to apply configuration to the sub device.
Hence, the application may get unwanted packets.
The fix adds error report for this case.
Fixes: ebea83f899d8 ("net/failsafe: add plug-in support")
C
The corrupted code didn't check the port availability when
it was trying to set the forward port IDs array.
However, when it was counting the number of ports, the availability
was checked by RTE_ETH_FOREACH_DEV iterator.
Hence, even when ETH devices ports were not in ATTACHED state,
the testpmd tr
This commit provides a set of tests for verifying the correctness and
performance of both unsigned 32 and 64bit reciprocal based division.
Signed-off-by: Pavan Nikhilesh
---
test/test/Makefile| 2 +
test/test/test_reciprocal_division.c | 109 ++
tes
Currently, rte_reciprocal only supports unsigned 32bit divisors. This
commit adds support for unsigned 64bit divisors.
Rename unsigned 32bit specific functions appropriately and update
librte_sched accordingly.
Signed-off-by: Pavan Nikhilesh
---
lib/librte_eal/bsdapp/eal/rte_eal_version.map |
In some use cases of integer division, denominator remains constant and
numerator varies. It is possible to optimize division for such specific
scenarios.
The librte_sched uses rte_reciprocal to optimize division so, moving it to
eal/common would allow other libraries and applications to use it.
28 matches
Mail list logo
