Stephen, what was the main reason you use red-black tree instead of dir-24-8?
Did you switch to using trees because of too big memory working set of
dir-24-8 algorithm?
2016-04-19 18:46 GMT+03:00 Stephen Hemminger :
> On Tue, 19 Apr 2016 14:11:11 +0300
> ? ??? wrote:
>
> > Hi.
> >
>
2016-04-29 17:34, Ferruh Yigit:
> Add default library output folder to the library search folder.
>
> This is useful for development environment, in production environment
> DPDK libraries already should be in know locations.
Yes it is useful in dev environment, but can be risky or strange when
p
Hi Jerin,
> -Original Message-
> From: Jerin Jacob [mailto:jerin.jacob at caviumnetworks.com]
> Sent: Monday, May 02, 2016 1:00 PM
> To: dev at dpdk.org
> Cc: De Lara Guarch, Pablo; Jerin Jacob
> Subject: [dpdk-dev] [PATCH] app/testpmd: add packet data prefetch in
> macswap loop
>
> prefe
2016-04-16 00:33, Jan Viktorin:
> The dpdk build system passes -mfloat-abi=softfp, which makes the build fail
> when the selected ABI is EABIhf. The dpdk build system should not make
> assumptions on the selected ARM ABI.
>
> Signed-off-by: Jan Viktorin
> Reported-by: Thomas Petazzoni
Applied,
the ifname[] field takes so much space, that it seperate some frequently
used fields into different caches, say, features and broadcast_rarp.
This patch move all those fields that will be accessed frequently in Rx/Tx
together (before the ifname[] field) to let them share one cache line.
Signed-of
Both current kernel virtio driver and DPDK virtio driver use at least
2 desc buffer for Tx: the first for storing the header, and the others
for storing the data.
Therefore, we could fetch the first data desc buf before the main loop,
and do the copy first before the check of "are we done yet?". T
Pre update and update used ring in batch for Tx and Rx at the stage
while fetching all avail desc idx. This would reduce some cache misses
and hence, increase the performance a bit.
Pre update would be feasible as guest driver will not start processing
those entries as far as we don't update "used
Here is a small patch set does the micro optimization, which brings about
10% performance boost in my 64B packet testing, with the following topo:
pkt generator <> NIC <-> Virtio NIC
Patch 1 pre updates the used ring and update them in batch. It should be
feasible from my understandin
Running with dpdk 16.04 and latest ovs from git, and removing "mrg_rxbuf=off"
from virtio params, the crash is no longer observed. However, we are
wittnessing ovs gets stuck, and will post to ovs mailing
list:2016-05-02T17:26:18.804Z|00111|ovs_rcu|WARN|blocked 1000 ms waiting for
pmd145 to quie
prefetch the next packet data address in advance in macswap loop
for performance improvement.
Signed-off-by: Jerin Jacob
---
app/test-pmd/macswap.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/app/test-pmd/macswap.c b/app/test-pmd/macswap.c
index 154889d..c10f4b5 100644
--- a/app/test-
Hi Daniel,
On 04/14/2016 03:01 PM, Daniel Mrzyglod wrote:
> This patch is for checking if error values occurs.
> fix for coverity errors #13209 & #13195
>
> If the function returns an error value, the error value may be mistaken
> for a normal value.
>
> In rdline_char_in: Value returned from a
2016-04-22 15:43, Thomas Monjalon:
> With this series, the default log level is not debug anymore.
> And more code depends on debug level instead of having some
> almost dead code.
>
> Thomas Monjalon (4):
> eal: increase log level of some messages
> log: increase default level to info
> exa
2016-04-29 16:29, Piotr Azarewicz:
> While parsing token string there may be several modes:
> - fixed single string
> - multi-choice single string
> - any single string
>
> This patch add one more mode - any multi string.
>
> Signed-off-by: Piotr Azarewicz
> Acked-by: Olivier Matz
Applied, tha
Hi Helin, thanks for the reply.
Some code might help me explain myself better-
port->configuration = rte_eth_conf {
.fdir_conf = {
.mode = RTE_FDIR_MODE_SIGNATURE,
.pballoc = RTE_FDIR_PBALLOC_64K,
.mask = rte_eth_fdir_masks {
From: Ilya Maximets
Array of buf_vector's is just an array for temporary storing information
about available descriptors. It used only locally in virtio_dev_merge_rx()
and there is no reason for that array to be shared.
Fix that by allocating local buf_vec inside virtio_dev_merge_rx().
Signed-o
Virtio net header length is set per device, but not per queue. So, there
is no reason to store it in vhost_virtqueue struct, instead, we should
store it in virtio_net struct, to make one copy only.
Signed-off-by: Yuanhan Liu
---
lib/librte_vhost/vhost-net.h | 2 +-
lib/librte_vhost/vhost_rxtx.
"virtio_net_device_ops" is the only left open struct that an application
can access, therefore, it's the only place that might introduce potential
ABI break in future for extension.
So, do some reservation for it. 5 should be pretty enough, considering
that we have barely touched it for a long whi
It barely has anything useful there, just 2 functions prototype. Here
move them to vhost-net.h, and delete it.
Signed-off-by: Yuanhan Liu
---
lib/librte_vhost/vhost-net.h | 3 ++
lib/librte_vhost/vhost_cuse/virtio-net-cdev.c | 1 -
lib/librte_vhost/vhost_rxtx.c
The "reserved" field in virtio_net and vhost_virtqueue struct is not
necessary any more. We now expose virtio_net device with a number "vid".
This patch also removes the "priv" field: all fields are priviate now:
application can't access it now. The only way that we could still access
it is to exp
We are now safe to move all those internal structs/macros/functions to
vhost-net.h, to hide them from external access.
This patch also breaks long lines and removes some redundant comments.
Signed-off-by: Yuanhan Liu
---
lib/librte_vhost/rte_virtio_net.h | 128 --
With all the previous prepare works, we are just one step away from
the final ABI refactoring. That is, to change current API to let them
stick to vid instead of the old virtio_net dev.
Signed-off-by: Yuanhan Liu
---
drivers/net/vhost/rte_eth_vhost.c | 61 ++-
Add few more functions to export few more fields or informations of
virtio_net struct, to applications, as we are gonna make them private.
It includes:
- rte_vhost_avail_entries
It's actually a rename of "rte_vring_available_entries", with the
"vring" to "vhost" name change to keep the consist
Query internal by vid instead of "ifname", to avoid the dependency of
virtio_net struct.
Signed-off-by: Yuanhan Liu
---
drivers/net/vhost/rte_eth_vhost.c | 19 +--
1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/drivers/net/vhost/rte_eth_vhost.c
b/drivers/net/vhos
vhost cuse is now the last reference of the vhost_device_ctx struct;
move it there, and do a rename to "vhost_cuse_device_ctx", to make
it clear that it's "cuse only".
Signed-off-by: Yuanhan Liu
---
lib/librte_vhost/vhost-net.h | 8
lib/librte_vhost/vhost_cuse/vhost-ne
get_device() just needs vid, so pass vid as the parameter only.
Signed-off-by: Yuanhan Liu
---
lib/librte_vhost/vhost-net.h | 30 ++--
lib/librte_vhost/vhost_cuse/vhost-net-cdev.c | 64 -
lib/librte_vhost/vhost_cuse/virtio-net-cdev.c | 18 ---
I failed to figure out what does "fh" mean here for a long while.
The only guess I could have had is "file handle". So, you get the
point that it's not well named.
I then figured it out that "fh" is derived from the fuse lib, and
my above guess is right. However, device_fh represents a virtio
net
Make a copy of virtio device id (device_fh) from the virtio_net struct,
so that we could have less dependency on the virtio_net struct.
Signed-off-by: Yuanhan Liu
---
examples/vhost/main.c | 59 ---
examples/vhost/main.h | 1 +
2 files changed, 29
Firstly, "int" would be big enough: we don't need 64 bit to represent
a virtio-net device id.
Secondly, this could let us avoid the ugly "%" PRIu64 ".." stuff.
And since ctx.fh is derived from device_fh, declare it as int, too.
Signed-off-by: Yuanhan Liu
---
examples/vhost/main.c
It does not make sense to ask the application to set/unset the flag
VIRTIO_DEV_RUNNING (that used internal only) at new_device()/
destroy_device() callback.
Instead, it should be set after new_device() succeeds and reset before
destroy_device() is invoked inside vhost lib. This patch fixes it.
Si
It's an fd; so define it as "int", which could also save the unncessary
(int) case.
Signed-off-by: Yuanhan Liu
---
lib/librte_vhost/rte_virtio_net.h | 2 +-
lib/librte_vhost/virtio-net.c | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/librte_vhost/rte_virtio_net.
Every time we introduce a new feature to vhost, we are likely
to break ABI. Moreover, some cleanups (such as the one from Ilya
to remove vec_buf from vhost_virtqueue struct) also break ABI.
This patch set is meant to resolve above issue ultimately, by
hiding virtio_net structure (as well as few ot
2016-04-27 17:07, Daniel Mrzyglod:
> Fix issue reported by clang scan-build
>
> Value of pointer tbl_next was uninitialized. When function lookup_step()
> take else branch it may provide garbage into tbl = tbl_next;
>
> Fixes: 5c510e13a9cb ("lpm: add IPv6 support")
>
> Signed-off-by: Daniel Mrzy
> > CID 13323:
> > Uninitialized scalar variable. Using uninitialized value
> > cfg->num_sections when calling rte_cfgfile_close.
> >
> > Fixes: eaafbad419bf ("cfgfile: library to interpret config files")
> >
> > Signed-off-by: Michal Kobylinski
>
> Acked-by: Cristian Dumitrescu
Applied, than
> > Function rte_cfgfile_load can return NULL value, when something goes
> > wrong.
> >
> > Signed-off-by: Dmitriy Yakovlev
> Acked-by: Cristian Dumitrescu
Applied, thanks
Embed dev_statistics into vhost_dev strcuct, which could clean
the code a bit.
Signed-off-by: Yuanhan Liu
---
examples/vhost/main.c | 87 +++
examples/vhost/main.h | 11 +--
2 files changed, 40 insertions(+), 58 deletions(-)
diff --git a/examp
switch_worker() is the last piece of code that is messy yet it touches
virtio/vhost device.
Here do a cleanup, so that we will be less painful for later vhost ABI
refactoring.
The cleanup is straigforward: break long lines, move some code into
functions. The last, comment a bit on switch_worker()
It has always been a mystery (at least to me before) that how many
mbuf is enough while creating an mbuf pool. While current macro
NUM_MBUFS_PER_PORT gives your some insights, it's not that accurate:
it doesn't consider the case we may receive a big packet, say 64K
when TSO is enabled.
We actually
Every time I do a VM2VM iperf test with vhost example, I have to set
the arp table manually, as vhost-switch just ignores the broadcast
packet, leaving the ARP request not served.
Here we do a transmit a broadcast packet (such as ARP request) to
every vhost device, as well as the physical port, to
rte_ether.h already provides a helper function to do mac address
compare. No need to define our own, use it directly.
Signed-off-by: Yuanhan Liu
---
examples/vhost/main.c | 14 +-
1 file changed, 1 insertion(+), 13 deletions(-)
diff --git a/examples/vhost/main.c b/examples/vhost/mai
To simplify code and logic.
Signed-off-by: Yuanhan Liu
---
examples/vhost/main.c | 457 --
examples/vhost/main.h | 32 ++--
2 files changed, 126 insertions(+), 363 deletions(-)
diff --git a/examples/vhost/main.c b/examples/vhost/main.c
index 9445
Interestingly, DESC_PER_CACHELINE has never been used since the
introduction of vhost example. Remove it.
vlan_ethhdr struct and VLAN_ETH_HLEN macro reference had been removed
by commit 4d50b6acbd95 ("examples/vhost: adapt Tx routing to lib"), but
had forgot to remove the definition.
Fixes: 4d50b
It's reported that it's has not been working for a long while. And due
to it's complex, it's better to redesign it than to fix it to make it
work again.
Signed-off-by: Yuanhan Liu
---
v2: remove macro PRINT_PACKET; will not be used anymore
---
doc/guides/sample_app_ug/vhost.rst | 36 +-
examp
I'm starting to work on the vhost ABI refactoring, that I also have to
touch the vhost example code. The vhost example code, however, is very
messy, full of __very__ long lines. This would make a later diff to
apply the new vhost API be very ugly, therefore, not friendly for review.
This is how thi
Elements of struct rte_eth_dev used in the fast path.
Make struct rte_eth_dev cache aligned to avoid the cases where
rte_eth_dev elements share the same cache line with other structures.
Signed-off-by: Jerin Jacob
---
lib/librte_ether/rte_ethdev.c | 2 +-
lib/librte_ether/rte_ethdev.h | 2 +-
2
This patch updates the parsing routines of packet queues (pktq_in/out
fields in the PIPELINE section) and message queues (msgq_in/out fields
of in the MSGQ Section) specified in ip_pipeline configuration file.
Signed-off-by: Jasvinder Singh
Acked-by: Cristian Dumitrescu
---
examples/ip_pipeline
2016-04-28 15:08, Igor Ryzhov:
> Include stdint.h for the definition of uint*_t types.
>
> Signed-off-by: Igor Ryzhov
Applied, thanks
> > Back then when we fixed the missing free lpm I was to quickly to say yes
> > if it applies not only to the lpm6 but also to all of the lpm code.
> >
> > It turned out to not apply to all of them. In rte_lpm_create_v20 there
> > is an unexpected fused allocation:
> > mem_size = sizeof(*lpm) + (s
On Thu, Apr 28, 2016 at 06:08:59PM +, Jianfeng Tan wrote:
> Issue: virtio's drv_flags are decided by devices types (modern vs legacy),
> and which kernel driver is used, and the negotiated features (especially
> VIRTIO_NET_STATUS) with backend, which makes it possible to multiple
> virtio devic
Hi Alex
Can you confirm that you are using DPDK? And how do you use DPDK and possibly
kernel driver?
I need your detailed topo of how are you using DPDK, as I am a bit confused.
Thanks!
Regards,
Helin
> -Original Message-
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Alex F
49 matches
Mail list logo
