[patch] mlx5: use after free in mlx5_cmd_comp_handler()

2013-07-22 Thread Dan Carpenter
We can't dereference ent after passing it to free_cmd(). Signed-off-by: Dan Carpenter dan.carpen...@oracle.com diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c index 205753a..4037406 100644 ---

Re: [PATCH for-3.11 7/7] IB/iser: Introduce fast memory registration model (FRWR)

2013-07-22 Thread Bart Van Assche
On 07/18/13 15:25, Or Gerlitz wrote: +static int iser_fast_reg_mr(struct fast_reg_descriptor *desc, + struct iser_conn *ib_conn, + struct iser_regd_buf *regd_buf, + u32 offset, unsigned int data_size, +

Re: [PATCH for-3.11 7/7] IB/iser: Introduce fast memory registration model (FRWR)

2013-07-22 Thread Sagi Grimberg
On 7/22/2013 2:46 PM, Bart Van Assche wrote: On 07/18/13 15:25, Or Gerlitz wrote: +static int iser_fast_reg_mr(struct fast_reg_descriptor *desc, +struct iser_conn *ib_conn, +struct iser_regd_buf *regd_buf, +u32 offset, unsigned int data_size, +

Re: [PATCH for-3.11 7/7] IB/iser: Introduce fast memory registration model (FRWR)

2013-07-22 Thread Or Gerlitz
On Mon, Jul 22, 2013 at 2:46 PM, Bart Van Assche bvanass...@acm.org wrote: [...] What will happen if a new FRWR is submitted with an rkey that is still valid ? The HW will fail this WR, we must invalidate the mapping before re-using the MR. Or. -- To unsubscribe from this list: send the line

[PATCH, RFC] mlx4: Avoid that mlx4_cmd_wait() contributes to the system load

2013-07-22 Thread Bart Van Assche
Avoid that kernel threads running mlx4_cmd_wait() contribute to the system load by setting the task state to TASK_INTERRUPTIBLE instead of TASK_UNINTERRUPTIBLE while waiting. This patch reduces the load average from about 0.5 to about 0.0 on an idle system with one mlx4 HCA and no IB cables

[PATCH 2/3] OpenSM: don't remove our own SM port.

2013-07-22 Thread Sasha Khapyorsky
When cleaning up remotes of non reachable ports anymore. Make exception for our own local SM port. In that case port still be in SM DB and running single port mode is possible. Signed-off-by: Sasha Khapyorsky sash...@gmail.com --- opensm/osm_drop_mgr.c | 4 +++- 1 file changed, 3 insertions(+),

Re: [PATCH] opensm: update internal PortInfo state for any ports

2013-07-22 Thread Sasha Khapyorsky
Hi Hal, On 09:37 Fri 12 Jul , Hal Rosenstock wrote: What difference does it make ? Does this patch fix some operational issue ? Is there any harm in not keeping the complete PortInfo when port is DOWN ? AFAIK it doesn't harm in the current state. However, I'm prepering the patches,

[PATCH 1/3] opensm: update internal PortInfo state for any ports

2013-07-22 Thread Sasha Khapyorsky
Should not be matter to keep internal SM's PortInfo data for ports in any states. Signed-off-by: Sasha Khapyorsky sash...@gmail.com --- opensm/osm_port.c | 33 - 1 file changed, 12 insertions(+), 21 deletions(-) diff --git a/opensm/osm_port.c b/opensm/osm_port.c

[PATCH 3/3] OpenSM: single port sweep

2013-07-22 Thread Sasha Khapyorsky
This provides possibility to keep SM/SA operational even in case when the local SM port was disconnected. It is needed in order to not break existing loopback connections. As side effect it let us to startup OpenSM on disconnected port. Signed-off-by: Sasha Khapyorsky sash...@gmail.com ---

[PATCH opensm] libvendor/osm_pkt_randomizer.c: Fix broken compilation with vendor sim

2013-07-22 Thread Hal Rosenstock
Due to change in number of args to osm_dr_path_init. Signed-off-by: Eitan Zahavi ei...@mellanox.com Signed-off-by: Alex Netes ale...@mellanox.com Signed-off-by: Hal Rosenstock h...@mellanox.com --- diff --git a/libvendor/osm_pkt_randomizer.c b/libvendor/osm_pkt_randomizer.c index