Re: [ewg] OFED-1.3.1 backport to RHEL 5 U 2 - beta

[Vladimir Sokolovsky]

Moshe Kazir wrote:
 
The attached patch file includes all the backport changes required for
 
OFED-1.3.1-20080520-0600.tgz back porting on RHEL 5 U2 beta kernel 
2.6.18-84.el5
 
Moshe 
 




Moshe Katzir   |  +972-9971-8639 (o)   |   +972-52-860-6042  (m)

 


Voltaire – _The Grid Backbone_

_ _

 www.voltaire.com 




Applied,
Thanks.

Moshe,
Can you upload 2.6.18-84.el5 sources to the OFA server?

Regards,
Vladimir
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] IB/ipoib: set max CM MTU when moving to CM mode

[Eli Cohen]

>From c878b9d3a4cfd031e8baaba46a224b46b1ced441 Mon Sep 17 00:00:00 2001
From: Eli Cohen <[EMAIL PROTECTED]>
Date: Thu, 22 May 2008 11:45:04 +0300
Subject: [PATCH] IB/ipoib: set max CM MTU when moving to CM mode

This will relieve the user from the need to restore CM mode MTU
every time we switch from UD to CM mode.

Signed-off-by: Eli Cohen <[EMAIL PROTECTED]>
---
I would like to push this patch to ofed 1.3.1 too.

 drivers/infiniband/ulp/ipoib/ipoib_cm.c |5 +
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c 
b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index 97e67d3..e6f57dd 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -1387,6 +1387,11 @@ static ssize_t set_mode(struct device *d, struct 
device_attribute *attr,
dev->features &= ~(NETIF_F_IP_CSUM | NETIF_F_SG | NETIF_F_TSO);
priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM;
 
+   if (ipoib_cm_max_mtu(dev) > priv->mcast_mtu)
+   ipoib_warn(priv, "mtu > %d will cause multicast packet 
drops.\n",
+  priv->mcast_mtu);
+   dev->mtu = ipoib_cm_max_mtu(dev);
+
ipoib_flush_paths(dev);
return count;
}
-- 
1.5.5.1



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] iSER

[Renaud Durand]

hello,

I tried to run iscsi on my SLES10 sp1 computers, for this, i followed the 
tutorials of OFED's wiki (https://wiki.openfabrics.org/tiki-index.php?page=iSER 
and https://wiki.openfabrics.org/tiki-index.php?page=ISER-target)
my problem is that : the target is working well (I can "discover" myself either 
with 127.0.0.1, my ethernet address and my ib address)

linux-target:~ # iscsiadm -m discovery -t sendtargets -p 127.0.0.1
161.74.X.X:3260,1 iqn.2001-04.com.example:storage.disk2.amiens.sys1.xyz
127.0.0.1:3260,1 iqn.2001-04.com.example:storage.disk2.amiens.sys1.xyz
linux-target:~ # 

but the remote computer can't discover the target

linux-cx5e:~ # iscsiadm -m discovery -t sendtargets -p 161.74.X.X
iscsiadm: connection to discovery address 161.74.X.X failed
iscsiadm: connection to discovery address 161.74.X.X failed
iscsiadm: connection to discovery address 161.74.X.X failed
iscsiadm: connection to discovery address 161.74.X.X failed
iscsiadm: connection to discovery address 161.74.X.X failed
iscsiadm: connection login retries (reopen_max) 5 exceeded
 
but the ping works 
linux-cx5e:~ # ping 161.74.X.X
PING 161.74.83.128 (161.74.X.X) 56(84) bytes of data.
64 bytes from 161.74.X.X: icmp_seq=1 ttl=64 time=2.27 ms
64 bytes from 161.74.X.X: icmp_seq=2 ttl=64 time=0.068 ms

here is lsmod on my computer

linux-cx5e:~ # lsmod | grep iscsi
iscsi_tcp  27520  0 
libiscsi   30208  1 iscsi_tcp
scsi_transport_iscsi34320  3 iscsi_tcp,libiscsi
scsi_mod  156600  6 
iscsi_tcp,libiscsi,scsi_transport_iscsi,sg,libata,sd_mod


I really don't understand what the problem is,
if you have a suggestion/solution please tell me because I am desperate 

_
Avec Hotmail, vos e-mails vous suivent partout ! Mettez Hotmail sur votre 
mobile !
http://www.messengersurvotremobile.com/?d=hotmail___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] OFED1.4 git

[Doron Shoham]

Hi,
Please revert commits 94845a467f29c6a4efb4c2b4cb658fb72d7aaca7 and 
e2e01970e31b201f065d0560db8dfbce1e978b70.
Those commits removed all of the iscsi/iser patches.
Next time we will appreciate consulting with us about patches that regards
to open-iscsi/iser issues.

Thanks,
Doron
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] OFED1.3.1 daily builds

[Doron Shoham]

Hi Vald,
I saw that in the latest daily builds didn't come with my commits
i.e on  open-iscsi-rh4.git tree commit num 
c79c9d187291f8845fcaa85b64de1838dad4f053.
I have sent you a patch which update the url open-iscsi gits (commit 
1fbe997c7a77f024483777060570e65fd94e924a)
Can it be that you are still pulling from Erez's old tree?

Thanks,
Doron
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] OFED1.3.1 daily builds

[Vladimir Sokolovsky]

Doron Shoham wrote:

Hi Vald,
I saw that in the latest daily builds didn't come with my commits
i.e on  open-iscsi-rh4.git tree commit num 
c79c9d187291f8845fcaa85b64de1838dad4f053.
I have sent you a patch which update the url open-iscsi gits (commit 
1fbe997c7a77f024483777060570e65fd94e924a)
Can it be that you are still pulling from Erez's old tree?

Thanks,
Doron


Hi Doron,
The problem is in your git tree:

You have both branch ofed_1_3 and tag ofed_1_3, so when the build_ofed.sh running 'git 
checkout -f ofed_1_3' it receives tag "ofed_1_3",
which is old (commit 93639f6d63518ec22527b7b915b0319813b7f50f) ...

To fix this, run: git-tag -d ofed_1_3

Please do it ASAP - I have to build OFED-1.3-rc2 now.

Regards,
Vladimir
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] OFED1.4 git

[Vladimir Sokolovsky]

Doron Shoham wrote:

Hi,
Please revert commits 94845a467f29c6a4efb4c2b4cb658fb72d7aaca7 and 
e2e01970e31b201f065d0560db8dfbce1e978b70.
Those commits removed all of the iscsi/iser patches.
Next time we will appreciate consulting with us about patches that regards
to open-iscsi/iser issues.

Thanks,
Doron


Hi Doron,
In order to create the kernel git tree for OFED-1.4 we used the following 
procedure:

1. Clone the kernel git tree of OFED-1.3 as ofed_1_4/linux-2.6.git
2. merge 2.6.26-rc2 kernel into ofed_1_4/linux-2.6.git
3. Remove all applied in 2.6.26-rc2 patches from kernel_patches/fixes directory
4. make ofed_scripts/ofed_makedist.sh to pass for all kernels.
   This step required to remove some patches (which were broken after the 
merge) from the
   kernel_patches/fixes directory.
   The removed patches should be recreated for 2.6.26-rc2 kernel and then they 
will be added
   to the kernel git tree of OFED-1.4
5. Create backport patches

So, please update your patches to fit 2.6.26-rc2 kernel and then update 
corresponding backport patches.

Regards,
Vladimir
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] update the open-iscsi rpm minor version

[Doron Shoham]

update open-iscsi rpm minor version from
2.0-754 to 2.0-754.1 and from 2.0-865.15 to 2.0-865.15.1

Signed-off-by: Doron Shoham <[EMAIL PROTECTED]>
---
 install.pl |4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/install.pl b/install.pl
index e608ba0..a795a3e 100755
--- a/install.pl
+++ b/install.pl
@@ -188,8 +188,8 @@ my @selected_by_user = ();
 my @selected_modules_by_user = ();
 my @selected_kernel_modules = ();
 
-my $open_iscsi_ver_rh4 = '2.0-754';
-my $open_iscsi_ver_non_rh4 = '2.0-865.15';
+my $open_iscsi_ver_rh4 = '2.0-754.1';
+my $open_iscsi_ver_non_rh4 = '2.0-865.15.1';
 
 sub usage
 {
-- 
1.5.3.8



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Britney and Paris lesbian video

[blayne jui-fen]

Watch 

quBjVVsgpD___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] IB/ipoib: copy small SKBs in CM mode

[Eli Cohen]

>From a8ea680caf189ad984aedaa81463ed66e45c4e65 Mon Sep 17 00:00:00 2001
From: Eli Cohen <[EMAIL PROTECTED]>
Date: Thu, 22 May 2008 16:28:59 +0300
Subject: [PATCH] IB/ipoib: copy small SKBs in CM mode

CM mode of ipoib has a large overhead in the receive flow for managing
SKBs. It usually allocates an SKB with data as much as was used in the
currently received SKB  and moves unused fragments from the old SKB to the
new one. This involves a loop on all the remaining fragments and incurs
overhead on the CPU.
This patch, for small SKBs, allocates an SKB just large enough to contain
the received data and copies to it the data from the received SKB.
The newly allocated SKB is passed to the stack and the old SKB is reposted.

Signed-off-by: Eli Cohen <[EMAIL PROTECTED]>
---

When running netperf I see significant improvement when using this patch
(BW Mbps):

with patch:
sender  receiver
313 313

without the patch:
509 134

 drivers/infiniband/ulp/ipoib/ipoib.h|1 +
 drivers/infiniband/ulp/ipoib/ipoib_cm.c |   15 +++
 2 files changed, 16 insertions(+), 0 deletions(-)

diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h 
b/drivers/infiniband/ulp/ipoib/ipoib.h
index ca126fc..e39bf36 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib.h
+++ b/drivers/infiniband/ulp/ipoib/ipoib.h
@@ -97,6 +97,7 @@ enum {
IPOIB_MCAST_FLAG_ATTACHED = 3,
 
MAX_SEND_CQE  = 16,
+   SKB_TSHOLD= 256,
 };
 
 #defineIPOIB_OP_RECV   (1ul << 31)
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c 
b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index e6f57dd..791bef7 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -525,6 +525,7 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct 
ib_wc *wc)
u64 mapping[IPOIB_CM_RX_SG];
int frags;
int has_srq;
+   struct sk_buff *small_skb;
 
ipoib_dbg_data(priv, "cm recv completion: id %d, status: %d\n",
   wr_id, wc->status);
@@ -579,6 +580,19 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct 
ib_wc *wc)
}
}
 
+   if (wc->byte_len < SKB_TSHOLD) {
+   int dlen = wc->byte_len;
+
+   small_skb = dev_alloc_skb(dlen + 12);
+   if (small_skb) {
+   skb_reserve(small_skb, 12);
+   skb_copy_from_linear_data(skb, small_skb->data, dlen);
+   skb_put(small_skb, dlen);
+   skb = small_skb;
+   goto copied;
+   }
+   }
+
frags = PAGE_ALIGN(wc->byte_len - min(wc->byte_len,
  (unsigned)IPOIB_CM_HEAD_SIZE)) / 
PAGE_SIZE;
 
@@ -601,6 +615,7 @@ void ipoib_cm_handle_rx_wc(struct net_device *dev, struct 
ib_wc *wc)
 
skb_put_frags(skb, IPOIB_CM_HEAD_SIZE, wc->byte_len, newskb);
 
+copied:
skb->protocol = ((struct ipoib_header *) skb->data)->proto;
skb_reset_mac_header(skb);
skb_pull(skb, IPOIB_ENCAP_LEN);
-- 
1.5.5.1



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH][TRIVIAL] ofed_scripts/90-ib.rules: uat has been deprecated for sometime now

[Hal Rosenstock]

ofed_scripts/90-ib.rules: uat has been deprecated for sometime now

Please apply to ofed_1_3 and ofed_1_4

Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]> 

diff --git a/ofed_scripts/90-ib.rules b/ofed_scripts/90-ib.rules
index 452cb98..994f4a0 100644
--- a/ofed_scripts/90-ib.rules
+++ b/ofed_scripts/90-ib.rules
@@ -2,6 +2,5 @@ KERNEL=="umad*", NAME="infiniband/%k"
 KERNEL=="issm*", NAME="infiniband/%k"
 KERNEL=="ucm*", NAME="infiniband/%k", MODE="0666"
 KERNEL=="uverbs*", NAME="infiniband/%k", MODE="0666"
-KERNEL=="uat", NAME="infiniband/%k", MODE="0666"
 KERNEL=="ucma", NAME="infiniband/%k", MODE="0666"
 KERNEL=="rdma_cm", NAME="infiniband/%k", MODE="0666"


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH][TRIVIAL] ofed_scripts/90-ib.rules: uat has been deprecated for sometime now

[Hal Rosenstock]

ofed_scripts/90-ib.rules: uat has been deprecated for sometime now

Please apply to ofed_1_3 and ofed_1_4

Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]> 

diff --git a/ofed_scripts/90-ib.rules b/ofed_scripts/90-ib.rules
index 452cb98..994f4a0 100644
--- a/ofed_scripts/90-ib.rules
+++ b/ofed_scripts/90-ib.rules
@@ -2,6 +2,5 @@ KERNEL=="umad*", NAME="infiniband/%k"
 KERNEL=="issm*", NAME="infiniband/%k"
 KERNEL=="ucm*", NAME="infiniband/%k", MODE="0666"
 KERNEL=="uverbs*", NAME="infiniband/%k", MODE="0666"
-KERNEL=="uat", NAME="infiniband/%k", MODE="0666"
 KERNEL=="ucma", NAME="infiniband/%k", MODE="0666"
 KERNEL=="rdma_cm", NAME="infiniband/%k", MODE="0666"


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] iSER

[Eli Dorfman]

On Thu, May 22, 2008 at 12:43 PM, Renaud Durand
<[EMAIL PROTECTED]> wrote:
> hello,
>
> I tried to run iscsi on my SLES10 sp1 computers, for this, i followed the
> tutorials of OFED's wiki
> (https://wiki.openfabrics.org/tiki-index.php?page=iSER and
> https://wiki.openfabrics.org/tiki-index.php?page=ISER-target)
> my problem is that : the target is working well (I can "discover" myself
> either with 127.0.0.1, my ethernet address and my ib address)
>
> linux-target:~ # iscsiadm -m discovery -t sendtargets -p 127.0.0.1
> 161.74.X.X:3260,1 iqn.2001-04.com.example:storage.disk2.amiens.sys1.xyz
> 127.0.0.1:3260,1 iqn.2001-04.com.example:storage.disk2.amiens.sys1.xyz
> linux-target:~ #
>
> but the remote computer can't discover the target
>
> linux-cx5e:~ # iscsiadm -m discovery -t sendtargets -p 161.74.X.X
> iscsiadm: connection to discovery address 161.74.X.X failed
> iscsiadm: connection to discovery address 161.74.X.X failed
> iscsiadm: connection to discovery address 161.74.X.X failed
> iscsiadm: connection to discovery address 161.74.X.X failed
> iscsiadm: connection to discovery address 161.74.X.X failed
> iscsiadm: connection login retries (reopen_max) 5 exceeded
>
> but the ping works
> linux-cx5e:~ # ping 161.74.X.X
> PING 161.74.83.128 (161.74.X.X) 56(84) bytes of data.
> 64 bytes from 161.74.X.X: icmp_seq=1 ttl=64 time=2.27 ms
> 64 bytes from 161.74.X.X: icmp_seq=2 ttl=64 time=0.068 ms
>
> here is lsmod on my computer
>
> linux-cx5e:~ # lsmod | grep iscsi
> iscsi_tcp  27520  0
> libiscsi   30208  1 iscsi_tcp
> scsi_transport_iscsi34320  3 iscsi_tcp,libiscsi
> scsi_mod  156600  6
> iscsi_tcp,libiscsi,scsi_transport_iscsi,sg,libata,sd_mod
>
>
> I really don't understand what the problem is,
> if you have a suggestion/solution please tell me because I am desperate
>

Please check the ACL in the target and verify that it includes the
initiator IB address
Run: tgtadm --lld iscsi --op show --mode target
If it is not you should add your initiator to the ACL using:
tgtadm --lld iscsi --mode target --op bind --tid=[id]
--initiator-address=[src-ip]

Also running iscsiadm -m discovery will create a node with tcp
transport and not iser.
If you want to use iser transport use iscsi_discovery script on the initiator.

Eli.
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] OFED 1.3.1 RC2 release is available

[Tziporet Koren]

Hi,

OFED 1.3.1 RC2 release is available on
http://www.openfabrics.org/downloads/OFED/ofed-1.3.1/OFED-1.3.1-rc2.tgz

To get BUILD_ID run ofed_info

Please report any issues in Bugzilla https://bugs.openfabrics.org/

The GA version is expected on May 29

Release information:

Linux Operating Systems:
- RedHat EL4 up4:   2.6.9-42.ELsmp
- RedHat EL4 up5:   2.6.9-55.ELsmp
- RedHat EL4 up6:   2.6.9-67.ELsmp
- RedHat EL5:   2.6.18-8.el5
- RedHat EL5 up1:   2.6.18-53.el5
- RedHat EL5 up2 beta:  2.6.18-84.el5   *
- Fedora C6:2.6.18-8.fc6*
- SLES10:   2.6.16.21-0.8-smp
- SLES10 SP1:   2.6.16.46-0.12-smp
- SLES10 SP1 up1:   2.6.16.53-0.16-smp
- SLES10 SP2:   2.6.16.60-0.21-smp  *
- OpenSuSE 10.3:2.6.22-*-*  *
- kernel.org:   2.6.23 and 2.6.24

  * OSes that are partially tested

Systems:
* x86_64
* x86
* ia64
* ppc64


Main changes from OFED 1.3.1-rc1

* Added backports for the OSes (with very limited testing):
* SLES10 SP2 with kernel 2.6.16.60-0.21-smp  
* RedHat EL5 up2 beta with kernel 2.6.18-84.el5   

* MPI packages update:
* mvapich-1.0.1-2481

* Updated libraries:
* dapl-v1 1.2.7-1
* dapl-v2 2.0.9-1
* libcxgb3 1.2.1

* ULPs changes:
   * OpenSM: Fix segmentation fault
   * iSER: Bug fixes since 2.6.24
   * RDS: fixes for RDMA API
   * IPoIB: Fix several kernel crashes (see attached list)

* Updated low level drivers:
   * nes
   * mlx4
   * cxgb3
   * ehca
   * ipath


Main Changes from OFED-1.3:
===
* MPI packages update:
* mvapich-1.0.1-2434
* mvapich2-1.0.3-1
* openmpi-1.2.6-1

* Updated libraries:
   * dapl-v1 1.2.6
   * dapl-v2 2.0.8
   * libcxgb3 1.2.0
   * librdmacm 1.0.7

* ULPs changes:
   * IB Bonding: ib-bonding-0.9.0-24
   * IPoIB bug fixes
   * RDS fixes for RDMA API
   * SRP failover

* Updated low level drivers:
   * nes
   * mlx4
   * cxgb3
   * ehca


Vlad & Tziporet

Note: 
In the attached tgz file you can find git-log of all changes.
In the CVS file there is a list of fixed bugs that were reported in
bugzilla

 <>  <> 


ofed-1.3.1_rc2-rc1.diff.tgz
Description: ofed-1.3.1_rc2-rc1.diff.tgz


rc2-fixed-bugs.csv
Description: rc2-fixed-bugs.csv
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg