Re: [net-next-2.6 PATCH] infiniband: convert to use netdev_for_each_mc_addr
Thu, Feb 25, 2010 at 09:00:07AM CET, ogerl...@voltaire.com wrote: Jiri Pirko wrote: --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -767,11 +767,8 @@ void ipoib_mcast_dev_flush(struct net_device *dev) -static int ipoib_mcast_addr_is_valid(const u8 *addr, unsigned int addrlen, - const u8 *broadcast) +static int ipoib_mcast_addr_is_valid(const u8 *addr, const u8 *broadcast) { -if (addrlen != INFINIBAND_ALEN) -return 0; This check was added by commit 5e47596b IPoIB: Check multicast address format, may I ask what is the reason for removing it now? Yes, at this very moment the check is not needless but it will be in a brief future. dev_mc_add will look very similar like dev_unicast_add. But ok. Here's patch adding the check in dev_mc_add right now to correct this state. Thanks Or. Subject: [net-next-2.6 PATCH] net: add addr len check to dev_mc_add Signed-off-by: Jiri Pirko jpi...@redhat.com diff --git a/net/core/dev_mcast.c b/net/core/dev_mcast.c index 9e2fa39..fd91569 100644 --- a/net/core/dev_mcast.c +++ b/net/core/dev_mcast.c @@ -96,6 +96,8 @@ int dev_mc_add(struct net_device *dev, void *addr, int alen, int glbl) int err; netif_addr_lock_bh(dev); + if (alen != dev-addr_len) + return -EINVAL; err = __dev_addr_add(dev-mc_list, dev-mc_count, addr, alen, glbl); if (!err) __dev_set_rx_mode(dev); -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
[ANNOUNCE] OFED 1.5.1 rc2 release is available
OFED 1.5.1-rc2 is available Notes: The tarball is available on: http://www.openfabrics.org/downloads/OFED/ofed-1.5.1/OFED-1.5.1-rc2.tgz To get BUILD_ID run ofed_info Please report any issues in bugzilla https://bugs.openfabrics.org/ for OFED 1.5.1 Vladimir Tziporet Supported Platforms and Operating Systems - o CPU architectures: - x86_64 - x86 - ppc64 - ia64 o Linux Operating Systems: - RedHat EL4 up72.6.9-78.ELsmp - RedHat EL4 up82.6.9-89.ELsmp - RedHat EL5 up32.6.18-128.el5 - RedHat EL5 up42.6.18-164.el5 - SLES10 SP22.6.16.60-0.21-smp - SLES10 SP32.6.16.60-0.54-smp - SLES112.6.27.19-5-default - OEL 4 up7 2.6.9-78.ELsmp - OEL 4 up8 2.6.9-89.ELsmp - CentOS5.3 2.6.18-128.el5 - CentOS5.4 2.6.18-164.el5 - Fedora Core12 2.6.31.5-127.fc12* - OpenSuSE 11.2 2.6.31.5-0.1-default * - kernel.org2.6.29, 2.6.30, 2.6.31 and 2.6.32* * Minimal QA for these versions Main changes from 1.5.1-rc1: === 1. Updated packages: - DAPL (compat-dapl-1.2.16, dapl-2.0.27) - ibutils: ibutils-1.5.2 - ib-bonding: ib-bonding-0.9.0-42 - libnes: libnes-1.0.0-0.1.g9b5f558 - libsdp: libsdp-1.1.100-0.1.g920ea31 - management: updated to the latest build - iSER: Added RHEL5.4 support - MPI: mvapich-1.2.0-3635 mvapich2-1.4-3.20100222svn3774 2. Bug fixes commit bb30f640fd973d48686e1663afbab67b23fa1c09 Merge: 03a9f17 329fbc0 Author: Vladimir Sokolovsky v...@mellanox.co.il Date: Wed Feb 24 08:47:38 2010 +0200 Merge branch 'code-drop/20100223' of git://git.openfabrics.org/~agrover/ofed_1_5/linux-2.6 into ofed_kernel_1_5 commit 03a9f17215dde8a49c8da22be88a8e1dc69ba754 Merge: cff66e5 19d717c Author: Vladimir Sokolovsky v...@mellanox.co.il Date: Wed Feb 24 08:40:15 2010 +0200 Merge branch 'ofed_kernel_1_5' of ssh://sofa.openfabrics.org/home/ctung/scm/ofed-1.5 into ofed_kernel_1_5 commit 329fbc005ecf6307377158aec7a7ce54c70777a9 Author: Tina Yang tina.y...@oracle.com Date: Fri Feb 19 16:53:00 2010 -0800 RDS: Fix locking in rds_send_drop_to() Signed-off-by: Tina Yang tina.y...@oracle.com Signed-off-by: Andy Grover andy.gro...@oracle.com commit 02955868562b143e4f81929275793bfa6c5b0ce7 Author: Andy Grover andy.gro...@oracle.com Date: Tue Feb 16 15:41:15 2010 -0800 RDS: Turn down alarming reconnect messages RDS's error messages when a connection goes down are a little extreme. A connection may go down, and it will be re-established, and everything is fine. This patch links these messages through rdsdebug(), instead of to printk directly. Signed-off-by: Andy Grover andy.gro...@oracle.com commit 3e7b1a2efceab7a4090fe8cbf5b8c2e2c67e7529 Author: Andy Grover andy.gro...@oracle.com Date: Tue Feb 16 13:54:12 2010 -0800 RDS: Workaround for in-use MRs on close causing crash if a machine is shut down without closing sockets properly, and freeing all MRs, then a BUG_ON will bring it down. This patch changes these to WARN_ONs -- leaking MRs is not fatal (although not ideal, and there is more work to do here for a proper fix.) Signed-off-by: Andy Grover andy.gro...@oracle.com commit 19d717c18ee40ca830bfed01839913329c548ed1 Author: Chien Tung chien.tin.t...@intel.com Date: Tue Feb 23 12:20:54 2010 -0600 RDMA/nes: add kernel_patches/fixes/nes_0027_kr.patch add support for KR device id 0x0110. nes_init_phy clean up. Deprecate NES_PHY_TYPE_IRIS. Signed-off-by: Chien Tung chien.tin.t...@intel.com commit cff66e5d6f3cb6dbd3ae68b7cbfd017a89833fa5 Merge: 464df47 51dbf00 Author: Vladimir Sokolovsky v...@mellanox.co.il Date: Tue Feb 23 17:38:33 2010 +0200 Merge remote branch 'amirv/ofed_kernel_1_5' into ofed_kernel_1_5 commit 51dbf00d8b656eaa89b95db86faa42f2106109c8 Author: Amir Vadai am...@mellanox.co.il Date: Tue Feb 23 10:02:59 2010 +0200 sdp: Prevent kernel crash if device init fails (plus bonus fix) If sdp_add_device() fails, there is no client data stored in the IB device, leading to a kernel crash when a connection is being established. Fix this by rejecting connections when the device is not initialized. Also, fix a bad goto target in an error case early in sdp_init_qp(). Signed-off-by: Joachim Fenkes fen...@de.ibm.com Signed-off-by: Amir Vadai am...@mellanox.co.il commit 464df477da4cb5996fdc39b95340b63797ed210b Author: Eli Cohen e...@mellanox.co.il Date: Tue Feb 23 14:23:55
Re: Oops in iscsi_conn_failure()
On 2010-02-25 02:06 Or Gerlitz said the following: Can you elaborate what does it take to reproduce that? I am not sure. I am writing custom scripts for a Linux cluster of iSCSI targets and initiators. I start and kill (including SIGKILL) iscsid, start and kill (including SIGKILL), add and remove targets using tgtadm, login and logout using iscsi_adm in some [random] order. Everything is done from the user space. Thanks. Roman -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: link width question / opensm
* Hal Rosenstock hal.rosenst...@gmail.com [2010-02-25 09:45:04]: The link won't renegotiate automatically. The link needs to be reset after updating the speed/width enabled values. Was that done ? I assumed OpenSM was resetting the link. Are you saying I need to run ibportstate reset in addition to running a modified opensm? (Or, using only ibportstate using ibportstate speed followed by a ibportstate reset?) - Is there another way to get what I want (control over link width+speed)? The latest ibportstate can do this. I installed the latest version, and it indeed has width and speed. However, it doesn't seem to have any effect: [dki...@f28]~% sudo /usr/sbin/ibportstate 232 1 query PortInfo: # Port info: Lid 232 port 1 LinkState:...Active PhysLinkState:...LinkUp LinkWidthSupported:..1X or 4X LinkWidthEnabled:1X or 4X LinkWidthActive:.4X LinkSpeedSupported:..2.5 Gbps or 5.0 Gbps or 10.0 Gbps LinkSpeedEnabled:2.5 Gbps or 5.0 Gbps or 10.0 Gbps LinkSpeedActive:.10.0 Gbps [dki...@f28]~% sudo /usr/sbin/ibportstate 232 1 speed 1 Initial PortInfo: # Port info: Lid 232 port 1 LinkSpeedEnabled:2.5 Gbps or 5.0 Gbps or 10.0 Gbps After PortInfo set: # Port info: Lid 232 port 1 LinkSpeedEnabled:2.5 Gbps [dki...@f28]~% sudo /usr/sbin/ibportstate 232 1 query PortInfo: # Port info: Lid 232 port 1 LinkState:...Active PhysLinkState:...LinkUp LinkWidthSupported:..1X or 4X LinkWidthEnabled:1X or 4X LinkWidthActive:.4X LinkSpeedSupported:..2.5 Gbps or 5.0 Gbps or 10.0 Gbps LinkSpeedEnabled:2.5 Gbps LinkSpeedActive:.10.0 Gbps [dki...@f28]~% sudo /usr/sbin/ibportstate 232 1 reset ibportstate: iberror: failed: smp query nodeinfo: Node type not switch Thanks, Dries pgpmlcOhhKkab.pgp Description: PGP signature
RE: [PATCH] RDMA/nes: add support for KR device id 0x0110
Iris is an XFP card. NetEffect didn't make that many and maybe a few were given out as evals. The biggest reason for removing the support is I don't have a card to test code changes. There are other cards that I am supporting even though we no longer make them as Intel. I will keep those around as long as I have cards, in a secret hiding place. :-) OK, if you're comfortable that no one is still using such cards then it's fine with me to remove the support. Thanks. Chien -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] IB/ipoib: fix dangling pointer references to ipoib_neigh and ipoib_path
On Thu, Feb 25, 2010 at 11:29:02AM -0800, Ralph Campbell wrote: I haven't looked carefully at the whole patch, but this bit looks wrong: @@ -848,61 +823,112 @@ static void ipoib_neigh_cleanup(struct neighbour *n) struct ipoib_neigh *neigh; struct ipoib_dev_priv *priv = netdev_priv(n-dev); unsigned long flags; - struct ipoib_ah *ah = NULL; + + spin_lock_irqsave(priv-lock, flags); neigh = *to_ipoib_neigh(n); - if (neigh) - priv = netdev_priv(neigh-dev); - else + if (neigh) { Should this be if (!neigh) ? + spin_unlock_irqrestore(priv-lock, flags); return; + } + *to_ipoib_neigh(n) = NULL; + neigh-neighbour = NULL; + -- Arthur -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: The SRP initiator and the endianness of tags in SRP information units
I have a question about the SRP initiator available in the Linux kernel. While consulting the SRP spec (r16a) I noticed that all multi-byte integer fields of SRP information units must be sent using the big endian byte order. This holds e.g. for the 64-bit tag fields present in several information units. So I was expecting these tags to be declared as __be64. However, in the header file scsi/srp.h these tags are declared as type u64, that is, being stored in CPU order. As a result, the byte order of the tags sent by the SRP initiator will depend on the endianness of the CPU the SRP initiator is running on (e.g. little endian on Intel CPU's, big endian on PowerPC CPU's). While this doesn't harm -- an SRP target must send back the same tag it received -- I found this confusing. Is this the intended behavior of the SRP initiator ? Didn't reply before... sorry. Anyway my view is that since tags are just an opaque 64-bit quantity, it doesn't really make sense to worry about endianness. You wouldn't byte-swap a field that was an 8-byte string, and the tag is essentially the same thing. - R. -- Roland Dreier rola...@cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/index.html -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Add new device IDs for ConnectX VPI HCAs
On Wed, Feb 24, 2010 at 02:07:20PM -0800, Roland Dreier wrote: + HCA(MELLANOX, 0x6746), /* MT26438 ConnectX VPI PCIe 2.0 5GT/s - IB QDR / 10GigE Virt+ */ The kernel calls this device ID MT26438 ConnectX EN 40GigE PCIe gen2 5GT/s. Maybe we should just delete these comments, since we can't seem to get them right? I guess it would be best if we take the description from http://pciids.sourceforge.net/v2.2/pci.ids + HCA(MELLANOX, 0x6778), /* MT26488 ConnectX VPI PCIe 2.0 5GT/s - IB DDR / 10GigE Virt+ */ Are we missing a kernel patch? I don't see this device ID in the kernel driver. Looks like it is missing. If we agree on this, we'll send two patches - one for kernel and one for libmlx4 to sync all device IDs with descriptions from the mentioned url. -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH] Add new device IDs for ConnectX VPI HCAs
The kernel calls this device ID MT26438 ConnectX EN 40GigE PCIe gen2 5GT/s. Maybe we should just delete these comments, since we can't seem to get them right? I guess it would be best if we take the description from http://pciids.sourceforge.net/v2.2/pci.ids I wonder whether it's really worth having the text at all. Maybe it's a bit useful, I'm not sure. But the pciids text could probably use some cleaning up, eg: 634a MT25418 [ConnectX VPI PCIe 2.0 2.5GT/s - IB DDR / 10GigE] 6368 MT25448 [ConnectX EN 10GigE, PCIe 2.0 2.5GT/s] it would be nice to decide on a standard order of whether the PCIe or the network info comes first. And then to make things even better we have: 103c 3313 HP NC542m Dual Port Flex-10 10GbE BLc Adapter so we can't even decide on 10GigE vs. 10GbE. But yeah, patches to get the ID list into shape with at least the right list of device IDs for kernel and userspace would be welcome. -- Roland Dreier rola...@cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/index.html -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH V2 2/2] mlx4/IB: Add support for enhanced atomic operations
| atomic_response = *va | if (((cmp XOR *va) AND cmp_mask) is ZERO) then | *va = (*va AND NOT(swap_mask)) OR (swap AND swap_mask) | | return atomic_response This is a great, terse description of the masked compare and swap operation. Do you think it would be more readable with C operations (ie ^ for XOR, for AND, ~ for NOT, and | for OR)? Also, I think it would be good to use the exact field names as the work request field you add (ie compare_add and compare_add_mask instead of cmp and cmp_mask). (And as I said before, this belongs with the generic patch, not the mlx4 patch description) Masked Fetch and Add (MFetchAdd) The MFetchAdd Atomic operation extends the functionality of the standard IB FetchAdd by allowing the user to split the target into multiple fields of selectable length. The atomic add is done independently on each one of this fields. A bit set in the field_boundary parameter specifies the field boundaries. The pseudo code below describes the operation: This is a bit harder to read, but I guess the operation is harder to describe. But what is really missing is the statement that the relevant fields in the wr are compare_add and compare_add_mask. Failing that, it's really hard to know how to use this operation without looking at the mlx4 source code. +props-device_cap_flags |= IB_DEVICE_MASKED_ATOMIC; Is there no firmware dependency here? If this really should be set unconditionally, then please just roll it into the other flags that get set for all devices. -- Roland Dreier rola...@cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/index.html -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
linux-next: manual merge of the infiniband tree with the vfs tree
Hi all, Today's linux-next merge of the infiniband tree got a conflict in drivers/infiniband/core/uverbs_main.c between commit ce916e2b935f8b3402da1457ff23b9f9f786c09b (switch infiniband uverbs to anon_inodes) from the vfs tree and commit 4169c4a9735d6434c9e39fa81ae5517e3afd4cd8 (IB/uverbs: Use anon_inodes instead of private infinibandeventfs) from the infiniband tree. These two commits purport to do something similar. Someone should look at them both and decide which one is right. For now I have used the version from the vfs tree - with the addition of the part from the infiniband tree version that selects ANON_INODES in the Kconfig file. -- Cheers, Stephen Rothwells...@canb.auug.org.au http://www.canb.auug.org.au/~sfr/ pgpTYxtornmxT.pgp Description: PGP signature