- 28 Sep, 2011 40 commits
-
-
Philipp Reisner authored
Signed-off-by:
Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by:
Lars Ellenberg <lars.ellenberg@linbit.com>
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
* drbd_update_congested() * we_should_drop_the_connection() Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
This reverts commit 6e9fdc92b77915d5c7ab8fea751f48378f8b0080. 1) This did not fixed the issue 2) Long sleeping work items can cause IO requests to take as long as the longest work item Signed-off-by:
-
Philipp Reisner authored
Work items that sleep too long can cause requests to take as long as the longest sleeping work item. Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
Describe how to reach any further overlapping intervals from the first overlap found. Signed-off-by:
-
Andreas Gruenbacher authored
Make it more clear that update_peer_seq() is supposed to wake up the seq_wait queue whenever the sequence number changes. Signed-off-by:
-
Andreas Gruenbacher authored
Before submitting a new local write request, wait for any conflicting local or remote requests to complete. We could assume that the new request occurred first and that the conflicting requests overwrote it (and therefore discard the new reques), but we know for sure that the new request occurred after the conflicting requests and so this behavior would we weird. We would also end up with the wrong result if the new request is not fully contained within the conflicting requests. Signed-off-by:
-
Andreas Gruenbacher authored
Nothing in this function accesses mdev->tconn->net_conf, so there is no need for get_net_conf() / put_net_conf() anymore. Signed-off-by:
-
Andreas Gruenbacher authored
Instead of keeping a separate tree for local and remote write requests for finding requests and for conflict detection, use the same tree for both purposes. Introduce a flag to allow distinguishing the two possible types of entries in this tree. Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
This flag is set when a processes puts itself to sleep to wait for a conflicting request to complete. Signed-off-by:
-
Andreas Gruenbacher authored
Move drbd_update_congested() to drbd_main.c, and drbd_req_new() and drbd_req_free() to drbd_req.c: those functions are not used anywhere else. Signed-off-by:
-
Andreas Gruenbacher authored
These things are only used there. Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
packet_seq is initialized by both sides of a connection in drbd_connect(). Signed-off-by:
-
Andreas Gruenbacher authored
Signed-off-by:
-
Andreas Gruenbacher authored
There is no good reason for cmdname() to be an inline function. Signed-off-by:
-
Philipp Reisner authored
Signed-off-by:
-
Philipp Reisner authored
Now drbd communication with protocol 100 actually works. Replaced the remaining p_header80 with p_header where we no longer know which header it is. In the places where p_header80 is still in use, it is on purpose, because we know that it is an old style header there. Signed-off-by:
-
