On 4/26/2017 3:18 PM, Laurence Oberman wrote:
----- Original Message -----
From: "Laurence Oberman" <loberman@xxxxxxxxxx>
To: "Max Gurtovoy" <maxg@xxxxxxxxxxxx>
Cc: "Leon Romanovsky" <leonro@xxxxxxxxxxxx>, "Bart Van Assche" <bart.vanassche@xxxxxxxxxxx>, "Doug Ledford"
<dledford@xxxxxxxxxx>, "Sagi Grimberg" <sagi@xxxxxxxxxxx>, "Israel Rukshin" <israelr@xxxxxxxxxxxx>,
linux-rdma@xxxxxxxxxxxxxxx
Sent: Wednesday, April 26, 2017 7:47:37 AM
Subject: Re: [PATCH, untested] mlx5: Avoid that mlx5_ib_sg_to_klms() overflows the klms[] array
----- Original Message -----
From: "Max Gurtovoy" <maxg@xxxxxxxxxxxx>
To: "Laurence Oberman" <loberman@xxxxxxxxxx>, "Leon Romanovsky"
<leonro@xxxxxxxxxxxx>
Cc: "Bart Van Assche" <bart.vanassche@xxxxxxxxxxx>, "Doug Ledford"
<dledford@xxxxxxxxxx>, "Sagi Grimberg"
<sagi@xxxxxxxxxxx>, "Israel Rukshin" <israelr@xxxxxxxxxxxx>,
linux-rdma@xxxxxxxxxxxxxxx
Sent: Wednesday, April 26, 2017 4:31:57 AM
Subject: Re: [PATCH, untested] mlx5: Avoid that mlx5_ib_sg_to_klms()
overflows the klms[] array
On 4/25/2017 11:37 PM, Laurence Oberman wrote:
----- Original Message -----
From: "Leon Romanovsky" <leonro@xxxxxxxxxxxx>
To: "Bart Van Assche" <bart.vanassche@xxxxxxxxxxx>
Cc: "Doug Ledford" <dledford@xxxxxxxxxx>, "Max Gurtovoy"
<maxg@xxxxxxxxxxxx>, "Sagi Grimberg" <sagi@xxxxxxxxxxx>,
"Israel Rukshin" <israelr@xxxxxxxxxxxx>, "Laurence Oberman"
<loberman@xxxxxxxxxx>, linux-rdma@xxxxxxxxxxxxxxx
Sent: Tuesday, April 25, 2017 1:58:49 PM
Subject: Re: [PATCH, untested] mlx5: Avoid that mlx5_ib_sg_to_klms()
overflows the klms[] array
On Mon, Apr 24, 2017 at 03:15:28PM -0700, Bart Van Assche wrote:
ib_map_mr_sg() can pass an SG-list to .map_mr_sg() that is larger
than what fits into a single MR. .map_mr_sg() must not attempt to
map more SG-list elements than what fits into a single MR.
Hence make sure that mlx5_ib_sg_to_klms() does not write outside
the MR klms[] array.
Fixes: b005d3164713 ("mlx5: Add arbitrary sg list support")
Signed-off-by: Bart Van Assche <bart.vanassche@xxxxxxxxxxx>
Reviewed-by: Max Gurtovoy <maxg@xxxxxxxxxxxx>
Cc: Sagi Grimberg <sagi@xxxxxxxxxxx>
Cc: Leon Romanovsky <leonro@xxxxxxxxxxxx>
Cc: Israel Rukshin <israelr@xxxxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
---
drivers/infiniband/hw/mlx5/mr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Bart,
Thanks a lot, it indeed looks right.
Acked-by: Leon Romanovsky <leonro@xxxxxxxxxxxx>
Thanks
Hello Bart, Leon, Max and Israel.
I cloned off Barts tree.
git clone https://github.com/bvanassche/linux
cd linux
git checkout block-scsi-for-next
I checked all patches were in for this test.
a83e404 IB/srp: Reenable IB_MR_TYPE_SG_GAPS
dfa5a2b mlx5: Avoid that mlx5_ib_sg_to_klms() overflows the klms[] array
f759c80 mlx5: Fix mlx5_ib_map_mr_sg mr lengt
Hi,
copying Sagi's request from different thread:
"
Can you please enable srp_add_one debug:
echo "func srp_add_one +p" > /sys/kernel/debug/dynamic_debug/control
In addition apply the following:
--
diff --git a/drivers/infiniband/hw/mlx5/mr.c
b/drivers/infiniband/hw/mlx5/mr.c
index d9c6c0ea750b..040fbc387e4f 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -1403,6 +1403,8 @@ mlx5_alloc_priv_descs(struct ib_device *device,
int add_size;
int ret;
+ WARN_ON_ONCE(ndescs > device->attr.max_fast_reg_page_list_len);
+
add_size = max_t(int, MLX5_UMR_ALIGN - ARCH_KMALLOC_MINALIGN, 0);
mr->descs_alloc = kzalloc(size + add_size, GFP_KERNEL);
"
Max.
Built and tested the kernel.
However this issue is not resolved :(
[ 2707.931909] scsi host1: ib_srp: failed RECV status WR flushed (5) for
CQE ffff8817edca86b0
[ 2708.089806] mlx5_0:dump_cqe:262:(pid 20129): dump error cqe
[ 2708.121342] 00000000 00000000 00000000 00000000
[ 2708.147104] 00000000 00000000 00000000 00000000
[ 2708.172633] 00000000 00000000 00000000 00000000
[ 2708.198702] 00000000 0f007806 2500002a 14a527d0
[ 2732.434127] scsi host1: ib_srp: reconnect succeeded
[ 2733.048023] scsi host1: ib_srp: failed RECV status WR flushed (5) for
CQE ffff8817ed0a9c30
[root@localhost ~]# [ 2746.413277] mlx5_0:dump_cqe:262:(pid 15877): dump
error cqe
[ 2746.443240] 00000000 00000000 00000000 00000000
[ 2746.469323] 00000000 00000000 00000000 00000000
[ 2746.495310] 00000000 00000000 00000000 00000000
[ 2746.521407] 00000000 0f007806 25000032 003c7ad0
[ 2752.445899] scsi host1: ib_srp: reconnect succeeded
[ 2752.481835] scsi host1: ib_srp: failed RECV status WR flushed (5) for
CQE ffff8817ed0a9cf0
[ 2763.267386] mlx5_0:dump_cqe:262:(pid 15877): dump error cqe
[ 2763.297826] 00000000 00000000 00000000 00000000
[ 2763.323352] 00000000 00000000 00000000 00000000
[ 2763.348722] 00000000 00000000 00000000 00000000
[ 2763.374681] 00000000 0f007806 2500003a 00084bd0
[root@localhost ~]# [ 2769.385203] fast_io_fail_tmo expired for SRP
port-1:1 / host1.
[ 2769.415956] scsi host1: ib_srp: reconnect succeeded
[ 2769.450258] scsi host1: ib_srp: failed RECV status WR flushed (5) for
CQE ffff8817ed0a9cf0
[ 2780.064627] mlx5_0:dump_cqe:262:(pid 18771): dump error cqe
[ 2780.093520] 00000000 00000000 00000000 00000000
[ 2780.120067] 00000000 00000000 00000000 00000000
[ 2780.145575] 00000000 00000000 00000000 00000000
[ 2780.171153] 00000000 0f007806 25000042 000833d0
[ 2785.923399] scsi host1: ib_srp: reconnect succeeded
[ 2785.957504] scsi host1: ib_srp: failed RECV status WR flushed (5) for
CQE ffff8817ed0a9cf0
[ 2796.463426] mlx5_0:dump_cqe:262:(pid 18771): dump error cqe
[ 2796.495257] 00000000 00000000 00000000 00000000
[ 2796.521506] 00000000 00000000 00000000 00000000
[ 2796.547640] 00000000 00000000 00000000 00000000
[ 2796.573120] 00000000 0f007806 2500004a 00083bd0
[ 2802.562578] scsi host1: ib_srp: reconnect succeeded
[ 2802.596880] scsi host1: ib_srp: failed RECV status WR flushed (5) for
CQE ffff8817ed0a9cf0
Regards
Laurence
Doing this now
Thanks
Laurence
Max
The Patch is not correct.
drivers/infiniband/hw/mlx5/mr.c: In function 'mlx5_alloc_priv_descs':
drivers/infiniband/hw/mlx5/mr.c:1406:30: error: 'struct ib_device' has no member named 'attr'
WARN_ON_ONCE(ndescs > device->attr.max_fast_reg_page_list_len);
^
./include/asm-generic/bug.h:117:27: note: in definition of macro 'WARN_ON_ONCE'
int __ret_warn_once = !!(condition); \
I think you meant to give me
WARN_ON_ONCE(ndescs > ib_device_attr->attr.max_fast_reg_page_list_len);
Can you confirm
Hi Laurence,
should be device->attrs.max_fast_reg_page_list_len.
please check this one that might solve the issue (on top of everything):
diff --git a/drivers/infiniband/hw/mlx5/mr.c
b/drivers/infiniband/hw/mlx5/mr.c
index b8f9382..063d116 100644
--- a/drivers/infiniband/hw/mlx5/mr.c
+++ b/drivers/infiniband/hw/mlx5/mr.c
@@ -1559,7 +1559,7 @@ struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd,
mr->max_descs = ndescs;
} else if (mr_type == IB_MR_TYPE_SG_GAPS) {
mr->access_mode = MLX5_MKC_ACCESS_MODE_KLMS;
-
+ MLX5_SET(mkc, mkc, translations_octword_size,
ALIGN(max_num_sg + 1, 4));
err = mlx5_alloc_priv_descs(pd->device, mr,
ndescs, sizeof(struct
mlx5_klm));
if (err)
thanks,
Max.
Thanks
Laurence
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
