Configuring UAR as IO-mapped makes maximum throughput decline by noticeable
amount. If UAR is configured as write-combining register, a write memory
barrier is needed on ringing a doorbell. rte_wmb() is mostly effective when
the size of a burst is comparatively small.
Personally I don't think
ff-by: Yongseok Koh
Signed-off-by: Alexander Solganik
Signed-off-by: Sagi Grimberg
---
drivers/net/mlx5/mlx5_rxtx.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/mlx5/mlx5_rxtx.h b/drivers/net/mlx5/mlx5_rxtx.h
index 7de1d10863e5..59b9ff24fb82 100644
--- a/drivers/net
When measuring latency when running a latency critical workload
on mlx5 pmd drivers we noticed high latency can occur due to
delayed doorbell record update flush.
This can be reproduced using the simple program [1]
against testpmd macswap fwd mode. This utility sends
a raw ethernet frame to the dp
doorbell after its write, setting
the mapping to un-cached by default.
Such change is reduces the max and average round trip
latency significantly.
Reported-by: Alexander Solganik
Signed-off-by: Shahaf Shuler
Signed-off-by: Yongseok Koh
Signed-off-by: Alexander Solganik
Signed-off-by: Sagi Grimberg
Looks good,
Reviewed-by: Sagi Grimberg
Looks good,
Reviewed-by: Sagi Grimberg
er small.
Yea, I don't mind merging it in given that I don't have time to come
up with anything better (or worse :))
Reviewed-by: Sagi Grimberg
Well, this is a fair argument, but without a *complete* solution for all
of dpdk peripherals, it has very little merit (if at all). A badly
written code can just as easily crash a server by passing a mbuf to
a crypto device or another network device that co-exists with mlx5.
So, while I underst
mlx5_tx_complete() polls completion queue multiple times until it
encounters an invalid entry. As Tx completions are suppressed by
MLX5_TX_COMP_THRESH, it is waste of cycles to expect multiple completions
in a poll. And freeing too many buffers in a call can cause high jitter.
This patch improves
I don't understand the security argument. Its completely private to the
driver. anything under librte is equivalent to an OS wrt networking, so I fail
to
see what is the security feature your talking about.
You are correct that as a root you are able to do whatever you want on the
server.
Th
mlx5_tx_complete() polls completion queue multiple times until it
encounters an invalid entry. As Tx completions are suppressed by
MLX5_TX_COMP_THRESH, it is waste of cycles to expect multiple completions
in a poll. And freeing too many buffers in a call can cause high jitter.
This patch improve
As I said, there are primitives which are designed to handle frequent reads
and rare mutations.
Even with such primitives, rarely lock is more than never lock.
You do realize that the cache mutation involves ibv_dereg_mr() right?
Any locking scheme for mutation is negligible compared to that
Its worse than just a drop, without debug enabled the error completion
is ignored, and the wqe_pi is taken from an invalid field, which leads
to bogus mbufs free (elts_tail is not valid).
Right
A simple work-around would be to simply fill a correct tail so that
error completions will still h
There is none, if you send a burst of 9 packets each one coming from a
different mempool the first one will be dropped.
Its worse than just a drop, without debug enabled the error completion
is ignored, and the wqe_pi is taken from an invalid field, which leads
to bogus mbufs free (elts_tail i
Hi,
Looking at the code, it looks like mlx5 keeps a MR cache per TX queue
(each MR registers a rte_mempool).
Once a TX queue is created, mlx5 scans existing mempools and
pre-registers a MR for each mempool it meets (using rte_mempool_walk).
For each MR registration that exceeds the TX queue cach
eeing it.
Fixes: a1bdb71a32da ("net/mlx5: fix crash in Rx")
Signed-off-by: Sagi Grimberg
Acked-by: Adrien Mazarguil
---
Changes from v2:
- fix check-git-log.sh complaints
- collected acked-by tag
Changes from v1:
- check pkt only once in case we failed to allocate a buffer
drivers/ne
eeing it.
Fixes: a1bdb71a32da ("net/mlx5: fix crash in Rx")
Signed-off-by: Sagi Grimberg
---
Changes from v1:
- check pkt only once in case we failed to allocate a buffer
drivers/net/mlx5/mlx5_rxtx.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/net/
On 02/08/16 12:58, Adrien Mazarguil wrote:
> On Tue, Aug 02, 2016 at 12:31:35PM +0300, Sagi Grimberg wrote:
>>
>>
>> On 01/08/16 19:43, Adrien Mazarguil wrote:
>>> Hi Sagi,
>>>
>>> On Mon, Aug 01, 2016 at 11:44:21AM +0300, Sagi Grimberg wrote:
On 01/08/16 19:43, Adrien Mazarguil wrote:
> Hi Sagi,
>
> On Mon, Aug 01, 2016 at 11:44:21AM +0300, Sagi Grimberg wrote:
>> The user is allowed to call ->rx_pkt_burst() even without free
>> mbufs in the pool. In this scenario we'll fail allocating a rep mbuf
>>
eeing it.
Signed-off-by: Sagi Grimberg
---
drivers/net/mlx5/mlx5_rxtx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/mlx5/mlx5_rxtx.c b/drivers/net/mlx5/mlx5_rxtx.c
index fce3381ae87a..a07cc4794023 100644
--- a/drivers/net/mlx5/mlx5_rxtx.c
+++ b/drivers/
20 matches
Mail list logo
