<feed xmlns='http://www.w3.org/2005/Atom'>
<title>kernel/linux.git/drivers/infiniband, branch v4.14.286</title>
<subtitle>Linux kernel stable tree (mirror)</subtitle>
<id>https://git.radix-linux.su/kernel/linux.git/atom?h=v4.14.286</id>
<link rel='self' href='https://git.radix-linux.su/kernel/linux.git/atom?h=v4.14.286'/>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/'/>
<updated>2022-06-14T14:53:55+00:00</updated>
<entry>
<title>RDMA/rxe: Generate a completion for unsupported/invalid opcode</title>
<updated>2022-06-14T14:53:55+00:00</updated>
<author>
<name>Xiao Yang</name>
<email>yangx.jy@fujitsu.com</email>
</author>
<published>2022-04-10T11:35:13+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=d798cacd13c5f3c74173e87b4462b8da3150558e'/>
<id>urn:sha1:d798cacd13c5f3c74173e87b4462b8da3150558e</id>
<content type='text'>
commit 2f917af777011c88e977b9b9a5d00b280d3a59ce upstream.

Current rxe_requester() doesn't generate a completion when processing an
unsupported/invalid opcode. If rxe driver doesn't support a new opcode
(e.g. RDMA Atomic Write) and RDMA library supports it, an application
using the new opcode can reproduce this issue. Fix the issue by calling
"goto err;".

Fixes: 8700e3e7c485 ("Soft RoCE driver")
Link: https://lore.kernel.org/r/20220410113513.27537-1-yangx.jy@fujitsu.com
Signed-off-by: Xiao Yang &lt;yangx.jy@fujitsu.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;
</content>
</entry>
<entry>
<title>RDMA/hfi1: Fix potential integer multiplication overflow errors</title>
<updated>2022-06-14T14:53:53+00:00</updated>
<author>
<name>Dennis Dalessandro</name>
<email>dennis.dalessandro@cornelisnetworks.com</email>
</author>
<published>2022-05-20T18:37:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=a89cb7ddf6a89bab6012e19da38b7cdb26175c19'/>
<id>urn:sha1:a89cb7ddf6a89bab6012e19da38b7cdb26175c19</id>
<content type='text'>
commit f93e91a0372c922c20d5bee260b0f43b4b8a1bee upstream.

When multiplying of different types, an overflow is possible even when
storing the result in a larger type. This is because the conversion is
done after the multiplication. So arithmetic overflow and thus in
incorrect value is possible.

Correct an instance of this in the inter packet delay calculation.  Fix by
ensuring one of the operands is u64 which will promote the other to u64 as
well ensuring no overflow.

Cc: stable@vger.kernel.org
Fixes: 7724105686e7 ("IB/hfi1: add driver files")
Link: https://lore.kernel.org/r/20220520183712.48973.29855.stgit@awfm-01.cornelisnetworks.com
Reviewed-by: Mike Marciniszyn &lt;mike.marciniszyn@cornelisnetworks.com&gt;
Signed-off-by: Dennis Dalessandro &lt;dennis.dalessandro@cornelisnetworks.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;
</content>
</entry>
<entry>
<title>RDMA/hfi1: Prevent panic when SDMA is disabled</title>
<updated>2022-06-14T14:53:47+00:00</updated>
<author>
<name>Douglas Miller</name>
<email>doug.miller@cornelisnetworks.com</email>
</author>
<published>2022-05-20T18:37:06+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=33794e8e9bcb4affc0ebff9cdec85acc8b8a1762'/>
<id>urn:sha1:33794e8e9bcb4affc0ebff9cdec85acc8b8a1762</id>
<content type='text'>
[ Upstream commit 629e052d0c98e46dde9f0824f0aa437f678d9b8f ]

If the hfi1 module is loaded with HFI1_CAP_SDMA off, a call to
hfi1_write_iter() will dereference a NULL pointer and panic. A typical
stack frame is:

  sdma_select_user_engine [hfi1]
  hfi1_user_sdma_process_request [hfi1]
  hfi1_write_iter [hfi1]
  do_iter_readv_writev
  do_iter_write
  vfs_writev
  do_writev
  do_syscall_64

The fix is to test for SDMA in hfi1_write_iter() and fail the I/O with
EINVAL.

Link: https://lore.kernel.org/r/20220520183706.48973.79803.stgit@awfm-01.cornelisnetworks.com
Signed-off-by: Douglas Miller &lt;doug.miller@cornelisnetworks.com&gt;
Signed-off-by: Dennis Dalessandro &lt;dennis.dalessandro@cornelisnetworks.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Sasha Levin &lt;sashal@kernel.org&gt;
</content>
</entry>
<entry>
<title>RDMA/ib_srp: Fix a deadlock</title>
<updated>2022-03-02T10:33:56+00:00</updated>
<author>
<name>Bart Van Assche</name>
<email>bvanassche@acm.org</email>
</author>
<published>2022-02-15T21:05:11+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=4752fafb461821f8c8581090c923ababba68c5bd'/>
<id>urn:sha1:4752fafb461821f8c8581090c923ababba68c5bd</id>
<content type='text'>
[ Upstream commit 081bdc9fe05bb23248f5effb6f811da3da4b8252 ]

Remove the flush_workqueue(system_long_wq) call since flushing
system_long_wq is deadlock-prone and since that call is redundant with a
preceding cancel_work_sync()

Link: https://lore.kernel.org/r/20220215210511.28303-3-bvanassche@acm.org
Fixes: ef6c49d87c34 ("IB/srp: Eliminate state SRP_TARGET_DEAD")
Reported-by: syzbot+831661966588c802aae9@syzkaller.appspotmail.com
Signed-off-by: Bart Van Assche &lt;bvanassche@acm.org&gt;
Reviewed-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Sasha Levin &lt;sashal@kernel.org&gt;
</content>
</entry>
<entry>
<title>RDMA/mlx4: Don't continue event handler after memory allocation failure</title>
<updated>2022-02-08T17:16:28+00:00</updated>
<author>
<name>Leon Romanovsky</name>
<email>leonro@nvidia.com</email>
</author>
<published>2022-01-31T09:45:26+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=cf7eb532dd6acbbbe1da5570fab659cc96d64148'/>
<id>urn:sha1:cf7eb532dd6acbbbe1da5570fab659cc96d64148</id>
<content type='text'>
commit f3136c4ce7acf64bee43135971ca52a880572e32 upstream.

The failure to allocate memory during MLX4_DEV_EVENT_PORT_MGMT_CHANGE
event handler will cause skip the assignment logic, but
ib_dispatch_event() will be called anyway.

Fix it by calling to return instead of break after memory allocation
failure.

Fixes: 00f5ce99dc6e ("mlx4: Use port management change event instead of smp_snoop")
Link: https://lore.kernel.org/r/12a0e83f18cfad4b5f62654f141e240d04915e10.1643622264.git.leonro@nvidia.com
Signed-off-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Reviewed-by: Håkon Bugge &lt;haakon.bugge@oracle.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;
</content>
</entry>
<entry>
<title>RDMA/rxe: Fix a typo in opcode name</title>
<updated>2022-01-27T08:01:00+00:00</updated>
<author>
<name>Chengguang Xu</name>
<email>cgxu519@mykernel.net</email>
</author>
<published>2021-12-18T11:23:20+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=6ba43e94d3e9c1ce499df7270493c2c7f819c854'/>
<id>urn:sha1:6ba43e94d3e9c1ce499df7270493c2c7f819c854</id>
<content type='text'>
commit 8d1cfb884e881efd69a3be4ef10772c71cb22216 upstream.

There is a redundant ']' in the name of opcode IB_OPCODE_RC_SEND_MIDDLE,
so just fix it.

Fixes: 8700e3e7c485 ("Soft RoCE driver")
Link: https://lore.kernel.org/r/20211218112320.3558770-1-cgxu519@mykernel.net
Signed-off-by: Chengguang Xu &lt;cgxu519@mykernel.net&gt;
Acked-by: Zhu Yanjun &lt;zyjzyj2000@gmail.com&gt;
Reviewed-by: Bob Pearson &lt;rpearsonhpe@gmail.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;
</content>
</entry>
<entry>
<title>RDMA/hns: Modify the mapping attribute of doorbell to device</title>
<updated>2022-01-27T08:01:00+00:00</updated>
<author>
<name>Yixing Liu</name>
<email>liuyixing1@huawei.com</email>
</author>
<published>2021-12-06T13:36:52+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=743a0aeb512e6423e06f80b1a5f140154053bc32'/>
<id>urn:sha1:743a0aeb512e6423e06f80b1a5f140154053bc32</id>
<content type='text'>
commit 39d5534b1302189c809e90641ffae8cbdc42a8fc upstream.

It is more general for ARM device drivers to use the device attribute to
map PCI BAR spaces.

Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver")
Link: https://lore.kernel.org/r/20211206133652.27476-1-liangwenpeng@huawei.com
Signed-off-by: Yixing Liu &lt;liuyixing1@huawei.com&gt;
Signed-off-by: Wenpeng Liang &lt;liangwenpeng@huawei.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Greg Kroah-Hartman &lt;gregkh@linuxfoundation.org&gt;
</content>
</entry>
<entry>
<title>RDMA/cxgb4: Set queue pair state when being queried</title>
<updated>2022-01-27T08:00:53+00:00</updated>
<author>
<name>Kamal Heib</name>
<email>kamalheib1@gmail.com</email>
</author>
<published>2021-12-20T15:25:30+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=001d3327cc77e85f3185fe0c1e793bba387fe2ab'/>
<id>urn:sha1:001d3327cc77e85f3185fe0c1e793bba387fe2ab</id>
<content type='text'>
[ Upstream commit e375b9c92985e409c4bb95dd43d34915ea7f5e28 ]

The API for ib_query_qp requires the driver to set cur_qp_state on return,
add the missing set.

Fixes: 67bbc05512d8 ("RDMA/cxgb4: Add query_qp support")
Link: https://lore.kernel.org/r/20211220152530.60399-1-kamalheib1@gmail.com
Signed-off-by: Kamal Heib &lt;kamalheib1@gmail.com&gt;
Reviewed-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Sasha Levin &lt;sashal@kernel.org&gt;
</content>
</entry>
<entry>
<title>RDMA/core: Let ib_find_gid() continue search even after empty entry</title>
<updated>2022-01-27T08:00:52+00:00</updated>
<author>
<name>Avihai Horon</name>
<email>avihaih@nvidia.com</email>
</author>
<published>2021-12-09T13:16:06+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=cb916e7da9a46779c174a0e1c3f5331c32373ce9'/>
<id>urn:sha1:cb916e7da9a46779c174a0e1c3f5331c32373ce9</id>
<content type='text'>
[ Upstream commit 483d805191a23191f8294bbf9b4e94836f5d92e4 ]

Currently, ib_find_gid() will stop searching after encountering the first
empty GID table entry. This behavior is wrong since neither IB nor RoCE
spec enforce tightly packed GID tables.

For example, when a valid GID entry exists at index N, and if a GID entry
is empty at index N-1, ib_find_gid() will fail to find the valid entry.

Fix it by making ib_find_gid() continue searching even after encountering
missing entries.

Fixes: 5eb620c81ce3 ("IB/core: Add helpers for uncached GID and P_Key searches")
Link: https://lore.kernel.org/r/e55d331b96cecfc2cf19803d16e7109ea966882d.1639055490.git.leonro@nvidia.com
Signed-off-by: Avihai Horon &lt;avihaih@nvidia.com&gt;
Reviewed-by: Mark Zhang &lt;markzhang@nvidia.com&gt;
Signed-off-by: Leon Romanovsky &lt;leonro@nvidia.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Sasha Levin &lt;sashal@kernel.org&gt;
</content>
</entry>
<entry>
<title>RDMA/hns: Validate the pkey index</title>
<updated>2022-01-27T08:00:52+00:00</updated>
<author>
<name>Kamal Heib</name>
<email>kamalheib1@gmail.com</email>
</author>
<published>2021-11-17T14:59:54+00:00</published>
<link rel='alternate' type='text/html' href='https://git.radix-linux.su/kernel/linux.git/commit/?id=3407ffb35c696b57d6e5f4a34cca52893a2f3798'/>
<id>urn:sha1:3407ffb35c696b57d6e5f4a34cca52893a2f3798</id>
<content type='text'>
[ Upstream commit 2a67fcfa0db6b4075515bd23497750849b88850f ]

Before query pkey, make sure that the queried index is valid.

Fixes: 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver")
Link: https://lore.kernel.org/r/20211117145954.123893-1-kamalheib1@gmail.com
Signed-off-by: Kamal Heib &lt;kamalheib1@gmail.com&gt;
Signed-off-by: Jason Gunthorpe &lt;jgg@nvidia.com&gt;
Signed-off-by: Sasha Levin &lt;sashal@kernel.org&gt;
</content>
</entry>
</feed>
