linux-toradex.git/drivers/infiniband/ulp/ipoib, branch v3.19.2

Revert "IPoIB: Consolidate rtnl_lock tasks in workqueue"

2015-01-30T23:39:36+00:00

This reverts commit afe1de664ef3cb756e70938d99417dcbc6b1379a.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: Make the carrier_on_task race aware"

2015-01-30T23:39:29+00:00

This reverts commit 67d7209e1f481cbaed37f9a224a328a3f83d0482.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: fix MCAST_FLAG_BUSY usage"

2015-01-30T23:39:20+00:00

This reverts commit 016d9fb25cd9817ea9c723f4f7ecd978636b4489.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: fix mcast_dev_flush/mcast_restart_task race"

2015-01-30T23:39:11+00:00

This reverts commit e5d1dcf1b0951f4ba00d93653942dda6196109d8.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: change init sequence ordering"

2015-01-30T23:39:02+00:00

This reverts commit 3bcce487fda8161597c20ed303d510e41ad7770e.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: Use dedicated workqueues per interface"

2015-01-30T23:38:55+00:00

This reverts commit 5141861cd5e17eac9676ff49c5abfafbea2b0e98.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: Make ipoib_mcast_stop_thread flush the workqueue"

2015-01-30T23:38:46+00:00

This reverts commit bb42a6dd02fb2901a69dbec2358810735b14b186.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

Revert "IPoIB: No longer use flush as a parameter"

2015-01-30T23:38:35+00:00

This reverts commit ce347ab90eaabc69a6146d41943981d51e7a9b82.

The series of IPoIB bug fixes that went into 3.19-rc1 introduce
regressions, and after trying to sort things out, we decided to revert
to 3.18's IPoIB driver and get things right for 3.20.

Signed-off-by: Roland Dreier

IPoIB: No longer use flush as a parameter

2014-12-16T02:11:15+00:00

Various places in the IPoIB code had a deadlock related to flushing
the ipoib workqueue.  Now that we have per device workqueues and a
specific flush workqueue, there is no longer a deadlock issue with
flushing the device specific workqueues and we can do so unilaterally.

Signed-off-by: Doug Ledford 
Signed-off-by: Roland Dreier

IPoIB: Make ipoib_mcast_stop_thread flush the workqueue

2014-12-16T02:11:15+00:00

We used to pass a flush variable to mcast_stop_thread to indicate if
we should flush the workqueue or not.  This was due to some code
trying to flush a workqueue that it was currently running on which is
a no-no.  Now that we have per-device work queues, and now that
ipoib_mcast_restart_task has taken the fact that it is queued on a
single thread workqueue with all of the ipoib_mcast_join_task's and
therefore has no need to stop the join task while it runs, we can do
away with the flush parameter and unilaterally flush always.

Signed-off-by: Doug Ledford 
Signed-off-by: Roland Dreier