The FIP proposal for VN_port to VN_port point-to-multipoint
operation requires a FLOGI be sent to each remote port.
The FLOGI is sent with the assigned S_ID and D_IDs of the
local and remote ports. This and the response get
FIP-encapsulated for Ethernet.
Add FLOGI state to the remote port state machine.
This will be skipped if not in point-to-multipoint mode.
To reduce a little duplication between PLOGI and FLOGI
response handling, added fc_rport_login_complete(), which
handles the parameters for the rdata struct.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
For VN_port to VN_port mode, the transport sets the port_id and
there's no lport FLOGI. This is similar to FC loop mode.
Add a point_to_multipoint flag that indicates the local port is in
point-to-multipoint mode. This skips FLOGI and discovery.
It also skips resetting the port_id on resets other than link down.
Add function fc_lport_set_local_id() that sets the local port_id.
This is called by libfcoe on behalf of the low-level driver
to set the port_id when the link comes up.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
For VN_port to VN_port mode, FIP will do discovery and needs a
way to find its state from the local port or discovery structure.
It seems that any other LLD that implements its own discovery
would also need something like this.
Replace disc->lport with disc->priv, and use container_of to
find the lport. We could use disc->priv for that, but
container_of is smaller and faster.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Add pre-zeroed space after the allocation for fc_rport_priv
for use by the lower-level driver.
This is primarily for VN2VN FIP mode, but could be used in
other ways someday.
The space required is specified in lport->rport_priv_size.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
To allow LLD to do lookups on rports without grabbing a mutex,
make them RCU-safe. The caller of lport->tt.rport_lookup will
have the choice of holding disc_mutex or the rcu_read_lock().
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
In this case, sync IO fails with EIO(5) errors as:-
"Thread:1 System call error:5 - Input/output error (::pwrite() failed)".
This is due to IO time out while libfc doing link down processing
to block all rports and if timed out IO was at last retry
attempt then it fails to user with EIO error followed by
these log messages.
[77848.612169] host2: rport bf0015: Delete port
[77848.612221] host2: rport e10aef: work delete
[77848.612232] host2: rport e10002: work event 3
[77848.612422] sd 2:0:1:1: [sdi] Unhandled error code
[77848.612426] sd 2:0:1:1: [sdi] Result: hostbyte=DID_ERROR
driverbyte=DRIVER_OK
[77848.612431] sd 2:0:1:1: [sdi] CDB: Write(10): 2a 00 00 00 11 20 00 00 20 00
[77848.612445] end_request: I/O error, dev sdi, sector 4384
[77848.612553] sd 2:0:1:2: [sdj] Unhandled error code
To fix these EIO errors, such timed out incomplete IOs needs
to be re-queued without counting retry attempt and this patch
does that using DID_REQUEUE scsi code.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This is exposed by a mpio test using EMC CLARiiON targets when LUN
tresspassing happens, the burst length from the XFER_READY for the
MODE SELECT(10) is 19 bytes, much smaller than FC_MIN_MAX_PAYLOAD as
256 bytes. This patch removes the related two WARN_ON()s.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Remote ports were restarting indefinitely after getting
rejects in PRLI.
Fix by adding a counter of restarts and limiting that with
the port login retry limit as well.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This patch somewhat combines two fixes to remote port handing in libfc.
The first problem was that rport work could be queued on a deleted
and freed rport. This is handled by not resetting rdata->event
ton NONE if the rdata is about to be deleted.
However, that fix led to the second problem, described by
Bhanu Gollapudi, as follows:
> Here is the sequence of events. T1 is first LOGO receive thread, T2 is
> fc_rport_work() scheduled by T1 and T3 is second LOGO receive thread and
> T4 is fc_rport_work scheduled by T3.
>
> 1. (T1)Received 1st LOGO in state Ready
> 2. (T1)Delete port & enter to RESTART state.
> 3. (T1)schdule event_work, since event is RPORT_EV_NONE.
> 4. (T1)set event = RPORT_EV_LOGO
> 5. (T1)Enter RESTART state as disc_id is set.
> 6. (T2)remember to PLOGI, and set event = RPORT_EV_NONE
> 6. (T3)Received 2nd LOGO
> 7. (T3)Delete Port & enter to RESTART state.
> 8. (T3)schedule event_work, since event is RPORT_EV_NONE.
> 9. (T3)Enter RESTART state as disc_id is set.
> 9. (T3)set event = RPORT_EV_LOGO
> 10.(T2)work restart, enter PLOGI state and issues PLOGI
> 11.(T4)Since state is not RESTART anymore, restart is not set, and the
> event is not reset to RPORT_EV_NONE. (current event is RPORT_EV_LOGO).
> 12. Now, PLOGI succeeds and fc_rport_enter_ready() will not schedule
> event_work, and hence the rport will never be created, eventually losing
> the target after dev_loss_tmo.
So, the problem here is that we were tracking the desire for
the rport be restarted by state RESTART, which was otherwise
equivalent to DELETE. A contributing factor is that we dropped
the lock between steps 6 and 10 in thread T2, which allows the
state to change, and we didn't completely re-evaluate then.
This is hopefully corrected by the following minor redesign:
Simplify the rport restart logic by making the decision to
restart after deleting the transport rport. That decision
is based on a new STARTED flag that indicates fc_rport_login()
has been called and fc_rport_logoff() has not been called
since then. This replaces the need for the RESTART state.
Only restart if the rdata is still in DELETED state
and only if it still has the STARTED flag set.
Also now, since we clear the event code much later in the
work thread, allow for the possibility that the rport may
have become READY again via incoming PLOGI, and if so,
queue another event to handle that.
In the problem scenario, the second LOGO received will
cause the LOGO event to occur again.
Reported-by: Bhanu Gollapudi <bprakash@broadcom.com>
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
lport state is enum not bit mask.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Resubmitting after incorporating Joe's review comment.
Unsolicited PRLO request is now handled by sending LS_ACC,
and then relogin to the remote port if an N-port login
session exists for that remote port.
Note that this patch should be applied on top of Joe Eykholt's
"Fix remote port restart problem" patch.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
As per FC-LS Rev 1.62 table 46, response codes are handled as follows:
1. If the Req executed is true, PRLI is accepted.
2. If Req executed is not set, if resp code is 5,
PRLI is not retried and port is logged out.
3. If resp code is anything apart from 1 or 5, PRLI is retired
upto max retry count.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Retry upto max_rport_retry_count when a target responds with
LS_RJT for a PRLI request.
Signed-off-by: Bhanu Prakash Gollapudi <bprakash@broadcom.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This patch creates a port_id member in struct fc_lport.
This allows libfc to just deal with fc_lport instances
instead of calling into the fc_host to get the port_id.
This change helps in only using symbols necessary for
operation from the libfc structures. libfc still needs
to change the fc_host_port_id() if the port_id changes
so the presentation layer (scsi_transport_fc) can provide
the user with the correct value, but libfc shouldn't
rely on the presentation layer for operational values.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
After the recent patch "fixes unnecessary seq id jump"
the SCST module fcst stopped working because multi-sequence
write data wasn't finding the sequence after the first frame.
Add back the setting of the seq_id when the first frame arrives.
Also fix indentation on two lines.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Print all world wide node names (node, port and fabric) with the same
format specifier of "%16.16llx". That makes sure they all print as a
16 character hex string, with lower case letters, no 0x prefix, and
without stripping off any leading 0s.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Most of the prints of fabric IDs were specified as %6x, which will not
print any leading 0s. It's nice to see leading 0s for identifiers
like this, which are a fixed length. This patch sets the precision
modifier as well, making the specifier %6.6x, which forces the
printing of leading 0s.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
fc_fcp_resp is assuming when FCP_SNS_LEN_VAL is set, the FCP_RSP_LEN_VAL
is not, which is not true. This leads to not copying the sense data and
error out a valid FCP_RSP.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
fc_exch_recv_req has variables eof, sof, and f_ctl,
which are set but never used. Delete them.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When the kernel is configured for preemption, using smp_processor_id()
when preemption is enabled causes a warning backtrace and is wrong
since we could move off of that CPU as soon as we get the ID,
and we would be referencing the wrong CPU, and possibly an invalid one
if it could be hotswapped out.
Remove the fc_lport_get_stats() function and explicitly use per_cpu_ptr()
to get the statistics. Where preemption has been disabled by holding
a _bh lock continue to use smp_processor_id(), but otherwise use
get_cpu()/put_cpu().
In fcoe_recv_frame() also changed the cases where we return in the
middle to do a goto to the code which bumps ErrorFrames and does
a put_cpu(). Two of these cases didn't bump ErrorFrames before, but
doing so is harmless because they "can't happen", due to prior length
checks.
Also rearranged code in fcoe_recv_frame() to have only one call to
fc_exch_recv(). It's just as efficient and saves a call to put_cpu().
In fc_fcp.c, adjusted a FIXME comment for code which doesn't need fixing.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
DID_ERROR cases can be ambigouos. Debugging FCP error cases
will be much easier if we have debug statements when we hit
these error conditions.
This patch simply adds debug messages using the FC_FCP_DBG
macro when we return DID_ERROR to SCSI. This way if a DID_ERROR
is reproducible turning on debug_logging will give a clue
to developers as to what the problem might be.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Currently fc_fcp_recv_data calls fc_fcp_retry_cmd to
retry failed IO but in this case tgt is still sending
data frames, therefore exchange needs to be aborted
first before initiating retry. So this patch fixes
this by aborting exchange first then have retry.
Renames fc_timeout_error to fc_fcp_recovery since
fc_timeout_error is already called from several other
places beside from fcp timeout handler and then
used fc_fcp_recovery for abort & retry from
fc_fcp_recv_data, this rename also required renaming
FC_CMD_TIME_OUT status to FC_CMD_RECOVERY to be
consistent with new fc_fcp_recovery.
Data frames are not expected for an DDPed exchange and
potentially it could be tampered data frame, so does
recovery in this case by calling fc_fcp_recovery.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Since use of offloads is more efficient than switching
to non-offload EM. However kept logic same to call em_match
if it is provided in the list of EMs.
Converted fc_exch_alloc to inline being now tiny a function
and already not an exported libfc API any more.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
In some cases seq is incremented twice causing unnecessary
seq jump, for instance fc_exch_recv_seq_resp increments
seq id when fc_sof_is_init is true and that is true for
each incoming xfer ready but then fc_fcp_send_data does
another seq increment to send data for xfer ready.
This patch removes all such seq id jumps, at least it
eliminates few calls to fc_seq_start_next using ex_lock.
Also removes seq id update with incoming frame's seq id
as this is not needed since each end (I or T) just need
to send incremented their own seq id on each TSI from
other end & before sending new sequence within a
exchange.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When starting a new response sequence in a multi-sequence
exchange, a warning was issued that sequence initiative
wasn't held.
The bug was that sequence initiative was cleared by the previous
sequence due to the END_SEQ flag being on. The intent may have
been to check LAST_SEQ. Change just to check SEQ_INIT.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The method we've been using for point-to-point mode requires
that the LS_ACC for the FLOGI uses the D_ID and S_ID assigned
to the remote port and local port, not those in the exchange.
This is not the correct method, but for now, it's what works
with the old target, as well as with new targets based on libfc.
This patch changes the addresses used accordingly.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When receiving a FLOGI request from a point-to-point peer,
the D_ID of 0xfffffe was not recognized as belonging to one
of the lports, so it was dropped.
Change fc_vport_id_lookup() to treat d_id 0xfffffe as a match.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
In point-to-point mode, if the PLOGI to the remote port times
out, it can get deleted by the remote port module. Since there's
no reference by the local port, lport->ptp_data points to a freed
rport, and when the local port is reset and tries to logout again,
an oops occurs in mutex_lock_nested().
Hold a reference count on the point-to-point rdata.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Reduce indentation in fc_rport_recv_prli_req() using gotos.
Also add payload length checks.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
Fix NULL pointer dereference crash occurs in fc_lport_bsg_request()
for bsg requests that do not contain a response request.
Specifically, FC_BSG_HST_ADD_RPORT and FC_BSG_HST_DEL_RPORT bsg
requests are not guaranteed to include a response request.
Signed-off-by: Hugh Daschbach <hdasch@broadcom.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Both PLOGI and RTV response processing conditionally scale e_d_tov,
but use different scaling factors. The scaling factor is correct in
RTV response processing. Bring PLOGI e_d_tov scaling in line with RTV
common service parameter inspection.
Signed-off-by: Hugh Daschbach <hdasch@broadcom.com>
Acked-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adds check to call fc_fcp_ddp_setup for only FCP read cmds to avoid
accessing junk fsp pointer at least in ESX since non FCP frame had
junk fsp value, though fsp is implicitly initialized to null
by __alloc_skb but with this patch no more relying on fsp
initialized to null value and hitting junk fsp ptr access.
Removes fsp pointer checking in fc_fcp_ddp_setup as this is not
needed any more since its only caller for FCP read will always
have a valid fsp.
Reported by: Frank Zhang <frank_1.zhang@intel.com>
Reported by: Rob Love <robert.w.love@intel.com>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
In particular, several occurances of funny versions of 'success',
'unknown', 'therefore', 'acknowledge', 'argument', 'achieve', 'address',
'beginning', 'desirable', 'separate' and 'necessary' are fixed.
Signed-off-by: Daniel Mack <daniel@caiaq.de>
Cc: Joe Perches <joe@perches.com>
Cc: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
This is to allow fcoemon util to enable or disable a fcoe interface
according to DCB link state change.
Adds sysfs module param enable and disable for this and also
updates existing other module param description to be consistent
and more accurate since older description had double "fcoe" word
with less meaningful netdev reference to user space.
Adds code to ignore redundant fc_lport_enter_reset handling for a
already disabled fcoe interface by checking LPORT_ST_DISABLED
or LPORT_ST_LOGO states, this also prevents lport state transition
on link flap on a disabled interface.
Above changes required lport state transition to get out of
disabled or logo state on call to fc_fabric_login.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Introduce a new lock to protect the list of fc_fcp_pkt structs in libfc
instead of using the host lock. This reduces the contention of this heavily
used lock, and I see up to a 25% performance gain in CPU bound small I/O
tests when scaling out across multiple quad-core CPUs.
The big win is in removing the host lock from the completion path
completely, as it does not need to be held around the call to scsi_done.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
We ran into a scenario where a remote port goes into RESTART state, but
never gets added to scsi transport. The running vmcore showed the following:
a) Port was in RESTART state
b) rdata->event was STOP
c) no work gets scheduled for the remote work to fc_rport_work
After this point, shut/no-shut of the remote port did not cause the port
to get re-discovered. The port would move betwen DELETE and RESTART states,
but the event would always be STOP, no work would get scheduled to
fc_rport_work and the port would not get added to scsi_transport.
The problem is that rdata->event is not set to NONE after a port is
restarted. After this point, no more work gets scheduled for the remote port
since new work is scheduled only if rdata->event is non-NONE. So, the event
and state keep changing, but fc_rport_work does not get scheduled to actually
handle the event.
Here's a transition of states that explains the above observation:
) Port is first in READY State, event is NONE
2) RSCN on shut, port goes to DELETED, event is stop
3) Before fc_rport_work runs, RSCN on no-shut, port goes to RESTART, event is
still STOP
4) fc_rport_work gets scheduled, removes the port from transport, sees state
as RESTART, begins the PLOGI state machine, event remains as STOP (event NOT
changed to NONE, this is the bug)
5) Plogi state machine completes, port state goes to READY, event goes to
READY, but no work is scheduled since event was STOP (non-NONE) before.
Fc_rport_work is not scheduled, port remains in READY state, but is not added
to transport.
Things are broken at this point. Libfc rport is ready, but no transport rport
created.
6) now a shut causes port state to change to DELETE, event to change to STOP,
no work gets scheduled
7) no-shut causes port state to change to RESTART, event remains at STOP,
no work gets scheduled
(6) and (7) now get repeated everytime we do shut/no-shut. No way to get out
of this state. Fcc reset does not help too.
Only way to get out is to load/unload module.
Fix is to set rdata->event to NONE while processing the STOP/LOGO/FAILED
events, inside the discovery and rport locks.
Signed-off-by: Abhijeet Joglekar <abjoglek@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Upon receiving ELS_RLS, send the Link Error Status Block (LESB) back.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Frame header room is already incluced, just pass the length of payload.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The FC-LS spec. says ELS timeouts should be 2 x R_A_TOV.
The FC-GS spec. says CT timeouts should be 3 x R_A_TOV.
We've been using E_D_TOV for both of those.
Change for all ELS and CT requests except FLOGI, which we
leave at 2 seconds (using E_D_TOV). One could argue that
R_A_TOV is locally determined until after FLOGI succeeds.
This does change FLOGI for vports which becomes FDISC.
This does not change the REC/SRR timeout which is 2 seconds.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
All exches must be freed before its EM mempool destroyed in this
case but currently some exches could be still pending in their
scheduled delayed work after EM mempool is destroyed causing
this issue discussed and reported in this latest email thread:-
http://www.open-fcoe.org/pipermail/devel/2009-October/004788.html
This patch fixes this issue by adding dedicated work queue thread
fc_exch_workqueue for exch delayed work and then flush this work
queue before destroying EM mempool.
The cancel_delayed_work_sync cannot be called during final
fc_exch_reset due to lport and exch locking ordering, so removes
related comment block not relevant any more with this patch.
Reported-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
There are cases outside of our control that may result in a transmit
skb being linearized in dev_queue_xmit. There are a couple of bugs
in libfc/fcoe that can result in a panic at that point. This patch
contains two fixes to prevent those panics.
1) use fast cloning instead of shared skbs with dev_queue_xmit
dev_queue_xmit doen't want shared skbuffs being passed in, and
__skb_linearize will BUG if the skb is shared. FCoE is holding an extra
reference around the call to dev_queue_xmit, so that when it returns an
error code indicating the frame has been dropped it can maintain it's
own backlog and retransmit. Switch to using fast skb cloning for this
instead.
2) don't append compound pages as > PAGE_SIZE skb fragments
fc_fcp_send_data will append pages from a scatterlist to the nr_frags[]
if the netdev supports it. But, it's using > PAGE_SIZE compound pages
as a single skb_frag. In the highmem linearize case that page will be
passed to kmap_atomic to get a mapping to copy out of, but
kmap_atomic will only allow access to the first PAGE_SIZE part.
The memcpy will keep going and cause a page fault once is crosses the
first boundary.
If fc_fcp_send_data uses linear buffers from the start, it calls
kmap_atomic one PAGE_SIZE at a time. That same logic needs to be
applied when setting up skb_frags.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
DID_NO_CONNECT is not a nice value to use for pkt alloc failures,
because you can probably retry and IO will become available again.
For the device reset callout, we do not want to set the scsi command
result for the above reason, and because we do not need to set
the scsi_cmd->result in this path. We and other drivers do not set it
for success for example, and we do not set it for other failure.
And scsi-ml does not send every command through this path, and it is
not expecting us to use the scsi_cmnd struct like a cmd coming thruogh
queuecommand. I think it is more for storage in case we need a cmd
struct for a tmf and to give us certain params like the LUN.
Patch was made over scsi-misc today.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Customers and certification tests have pointed out that we don't
show up on the switch management software as an initiator.
On some MDS switches 'show fcns database' command shows libfc
initiators as 'fcp' not 'fcp:init' like other initiators.
On others switches, I think the switch gets the features by doing a PRLI,
but it may be only certain models or under certain configurations.
Fix this by registering our FC4 features with the RFF_ID CT request
after local port login and after the RFT_ID.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The libfc link up/down messages don't indicate which port is changing.
The Port ID will often be 0.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This is to notify the LLD when an FC_ID is assigned to the local port.
The fnic driver needs to push the assigned FC_ID to firmware.
It currently does this by intercepting the FLOGI responses, and
in order to make that code more common with FIP and NPIV, it
makes more sense to wait until the local port has completely
handled the FLOGI or FDISC response. Also, when we fix
point-to-point FC_ID assignment, we'll need this callback as well.
Add a call to the libfc template, which is called whenever
the local port FC_ID is being assigned. It defaults to
fc_lport_set_fid(), supplied by libfc.
As additional benefit of this function, the LLD may determine
the MAC address that caused the change by looking at the received frame.
We also print the assigned port ID as long as it isn't 0.
Setting port ID to 0 happens often in reset while retrying FLOGI,
and would be uninteresting. This replaces the previous message
which didn't identify the host adapter instance.
patch v2 note: changed one word in a comment. "intercepted" -> "provided".
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Local port debug messages were using fc_els_resp_type() which showed
all CT responses as rejects.
Handle CT responses correctly based by inspecting fh_type.
I decided not to rename the function to keep the patch smaller.
We could call it just fc_resp_type() or fc_elsct_resp_type().
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adds last_can_queue_ramp_down_time and updates this on every
ramp down. If last_can_queue_ramp_down_time is not zero then
do ramp up on any IO completion in added fc_fcp_can_queue_ramp_up.
Reset last_can_queue_ramp_down_time to zero once can_queue
is ramped up to added max_can_queue limit, this is to avoid any
more ramp up attempts on subsequent IO completion.
The ramp down and up are skipped for FC_CAN_QUEUE_PERIOD
to avoid infrequent changes to can_queue, this required
keeping track of ramp up time also in last_can_queue_ramp_up_time.
Adds code to ramp down can_queue if lp->qfull is set, with added
new ramp up code the can_queue will be increased after
FC_CAN_QUEUE_PERIOD, therefore it is safe to do ramp down
without fsp in this case and will avoid thrash. This required
fc_fcp_can_queue_ramp_down locking change so that it can be
called with Scsi_Host lock held.
Removes si->throttled and fsp state FC_SRB_NOMEM, not needed with
added ramp up code.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Currently can_queue is reduced only if frame alloc fails
during fc_fcp_send_data but frame alloc can fail at several
other places in FCP data path and can_queue needs to be
reduced for any FCP frame alloc failure.
This patch adds fc_fcp_frame_alloc for all FCP frame allocations
and if fc_frame_alloc fails in fc_fcp_frame_alloc then reduce
can_queue in fc_fcp_frame_alloc, this will reduce can_queue for
all FCP frame alloc failures.
This required moving fc_fcp_reduce_can_queue up, to build without
adding its prototype. Also renamed fc_fcp_reduce_can_queue to
fc_fcp_can_queue_ramp_down.
Removes fc_fcp_reduce_can_queue calling from fc_fcp_recv since
not needed with added fc_fcp_frame_alloc reducing can_queue.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Cleans up frame allocation APIs to have just single fc_frame_alloc API.
Removes _fc_frame_alloc, renames __fc_frame_alloc to _fc_frame_alloc.
Modifies fc_fcp_send_data for removed _fc_frame_alloc, fc_fcp_send_data
was the only user of removed _fc_frame_alloc.
Also Adds check in fc_frame_alloc to do mod by 4 for only non-zero
len value.
This patch is prep work to fix can_queue reducing in next patch.
Single fc_frame_alloc API helps in fixing can_queue reducing in
next patch.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This patch makes a variety of cleanup changes to all libfc files.
This patch adds kernel-doc headers to all functions lacking them
and attempts to better format existing headers. It also add kernel-doc
headers to structures.
This patch ensures that the current naming conventions for local ports,
remote ports and remote port private data is upheld in the following
manner.
struct instance (i.e. variable name)
--------------------------------------------------
fc_lport lport
fc_rport rport
fc_rport_libfc_priv rpriv
fc_rport_priv rdata
I also renamed dns_rp and ptp_rp to dns_rdata and ptp_rdata
respectively.
I used emacs 'indent-region' and 'tabify' on all libfc files
to correct spacing alignments.
I feel sorry for anyone attempting to review this patch.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This is the Open-FCoE implementation of the FC
passthrough support via bsg interface.
Passthrough support is added to both N_Ports and
VN_Ports.
Signed-off-by: Steve Ma <steve.ma@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When handling the multi-frame responses of fc pass-thru requests,
a code segment similar to fc_fcp_recv_data (routine to receive
inbound SCSI data) is used in the response handler. This patch
is to add a routine, called fc_copy_buffer_to_sglist(), to handle
the common function of copying data from a buffer to a scatter-
gather list in order to avoid code duplication.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Allow a vport specific string to be appended to the port symbolic
name. The new symbolic name is sent to the name server after it
is set.
This currently messes with libhbalinux, which is looking for
the fcoe "fcoe <ver> over <ethX>" string and expects whatever
comes after the "over" to be a network interface name only.
Adds an EXPORT_SYMBOL to libfc for fc_frame_alloc_fill, which is
needed to allow fcoe to allocate a frame of variable length for
the RSPN request.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
They all do the same thing, so combine them into a single function.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Register the fc_host symbolic name as the symbolic port name
with the fabric name server.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Register the fc_host symbolic name as the symbolic node name
with the fabric name server.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
One could interpret FC-GS-5 to say that an explicit RNN_ID is required
before RSNN_NN is allowed to succeed, which is why RNN_ID was not obsoleted
along with RPN_ID acording to this document:
ftp://ftp.t11.org/t11/member/fc/gs-5/05-546v2.pdf
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
RPN_ID has been obsolete per FC-GS-5 for several years. The port name is
registered implicitly as part of FLOGI, and it is undesirable for ports to
change a registered port name using RPN_ID while logged into the fabric.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The FIP code in libfcoe needed several changes to support NPIV
1) dst_src_addr needs to be managed per-n_port-ID for FPMA fabrics with NPIV
enabled. Managing the MAC address is now handled in fcoe, with some slight
changes to update_mac() and a new get_src_addr() function pointer.
2) The libfc elsct_send() hook is used to setup FCoE specific response
handlers for FIP encapsulated ELS exchanges. This lets the FCoE specific
handling know which VN_Port the exchange is for, and doesn't require
tracking OX_IDs. It might be possible to roll back to the full FIP frame
in these, but for now I've just stashed the contents of the MAC address
descriptor in the skb context block for later use. Also, because
fcoe_elsct_send() just passes control on to fc_elsct_send(), all transmits
still come through the normal frame_send() path.
3) The NPIV changes added a mutex hold in the keep alive sending, the lport
mutex is protecting the vport list. We can't take a mutex from a timer,
so move the FIP keep alive logic to the link work struct.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Add FDISC ELS handling to libfc and libfcoe, treat it the same as FLOGI where
appropriate.
Add checking for NPIV support in the FLOGI LS_ACC service parameters.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
NPIV vports are managed in libfc by changing their virtual link state
when the parent N_Ports internal state changes. The vport link is only
online when the N_Port is in a ready state (logged into the fabric).
vport_state is updated as needed in this patch as well, currently the states
LINKDOWN, INITIALIZING, ACTIVE, DSIABLED, and NO_FABRIC_SUPP are used.
This also changes the fc_host port_state handling to differentiate between
LINKDOWN and OFFLINE.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adds a function to create a new VN_Port instances, which share the EM
list with the N_Port, VN_Port lookup by fabric ID when responding to a new
request (otherwise the exchange lookup from the N_Ports EM list is trusted to
return an exchange with a cached lport value for the correct VN_Port),
a pointer to a fc_vport structure for VN_Ports, and flags to indicate if an
N_Port supports NPIV and if the switch/fabric allows it.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
I'd like to keep basic initialization together with allocation, which means
this can't just be a tail-call to scsi_host_alloc.
This is needed to create a generic libfc host allocation routine for NPIV
VN_Ports, which will share the exchange ID space (through sharing exchange
manager structures) with the parent lport. In order to clone the exchange
manager list when the lport is allocated, the list head must be initialized
earlier.
Also, update fnic to use the libfc_host_alloc so that later changes do not break
it. (contribution by Joe Eykholt)
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
These routines are for the libfc kernel module and should be in
the libfc .c file.
Moving the libfc __init routine into fc_libfc.c caused the creation
of the fc_setup_fcp() and fc_destroy_fcp() routines so that
scsi_pkt_cachep was not exposed outside of fc_fcp.c.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
include/scsi/libfc.h is currently loaded with common code
shared between libfc's sub-modules as well as shared between
libfc and fcoe. Previous patches attempted to move out
non-common code. This patch creates two files for common
libfc routines that will not be shared with fcoe, fnic or
any other LLDs.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This function is never used, let's remove it.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This patch moves all non-common routines and function prototypes
out of libfc.h and into the appropriate .c files. It makes these
routines 'static' when necessary and removes any unnecessary EXPORT_SYMBOL
statements.
A result of moving the fc_exch_seq_send, fc_seq_els_rsp_send, fc_exch_alloc
and fc_seq_start_next prototypes out of libfc.h is that they were no longer
being imported into fc_exch.c when libfc.h was included. This caused errors
where routines in fc_exch.c were looking for undefined symbols. To fix this
this patch reorganizes fc_seq_alloc, fc_seq_start_next and
fc_seq_start_next_locked. This move also made it so that
fc_seq_start_next_locked did not need to be prototyped at the top of
fc_exch.c.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adjust queue_depth on fc_change_queue_depth call back
with reason SCSI_QDEPTH_RAMP_UP, no additional resource
adjustments necessary for libfc.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Acked-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The cmd_per_lun value is used by scsi-ml as fall back lowest
queue_depth value but in case of libfc cmd_per_lun is set to
same value as max queue_depth = 32.
So this patch reduces cmd_per_lun value to 3 and configures
each lun with default max queue_depth 32 in fc_slave_alloc.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Acked-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This converts the libfc using scsi_track_queue_full to
track the queue full from the change_queue_depth callback.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Acked-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This patch modifies scsi_host_template->change_queue_depth so that
it takes an argument indicating why it is being called. This will be
used so that if a LLD needs to do some extra processing when
handling queue fulls or later ramp ups, it can do so.
This is a simple port of the drivers setting a change_queue_depth
callback. In the patch I just have these LLDs adjust the queue depth
if the user was requesting it.
Signed-off-by: Mike Christie <michaelc@cs.wisc.edu>
[Vasu.Dev: v2
Also converted pmcraid_change_queue_depth and then verified
all modules compile using "make allmodconfig" for any new build
warnings on X86_64.
Updated original description after combing two original
patches from Mike to make this patch git bisectable.]
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
[jejb: fixed up 53c700]
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Timer crashes were caused by freeing a struct fc_rport_priv
with a timer pending, causing the timer facility list to be
corrupted. This was during FC uplink flap tests with a lot
of targets.
After discovery, we were doing an PLOGI on an rdata that was
in DELETE state but not yet removed from the lookup list.
This moved the rdata from DELETE state to PLOGI state.
If the PLOGI exchange allocation failed and needed to be
retried, the timer scheduling could race with the free
being done by fc_rport_work().
When fc_rport_login() is called on a rport in DELETE state,
move it to a new state RESTART. In fc_rport_work, when
handling a LOGO, STOPPED or FAILED event, look for restart
state. In the RESTART case, don't take the rdata off the
list and after the transport remote port is deleted and
exchanges are reset, re-login to the remote port.
Note that the new RESTART state also corrects a problem we
had when re-discovering a port that had moved to DELETE state.
In that case, a new rdata was created, but the old rdata
would do an exchange manager reset affecting the FC_ID
for both the new rdata and old rdata. With the new state,
the new port isn't logged into until after any old exchanges
are reset.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
I was running into several different panics under stress, which I traced down
to a few different possible slab corruption issues in error handling paths.
I have not yet looked into why these exchange sends fail, but with these
fixes my test system is much more stable under stress than before.
fc_elsct_send() could fail and either leave the passed in frame intact
(failure in fc_ct/els_fill) or the frame could have been freed if the
failure was is fc_exch_seq_send(). The caller had no way of knowing, and
there was a potential double free in the error handling in fc_fcp_rec().
Make fc_elsct_send() always free the frame before returning, and remove the
fc_frame_free() call in fc_fcp_rec().
While fc_exch_seq_send() did always consume the frame, there were double free
bugs in the error handling of fc_fcp_cmd_send() and fc_fcp_srr() as well.
Numerous calls to error handling routines (fc_disc_error(),
fc_lport_error(), fc_rport_error_retry() ) were passing in a frame pointer that
had already been freed in the case of an error. I have changed the call
sites to pass in a NULL pointer, but there may be more appropriate error
codes to use.
Question: Why do these error routines take a frame pointer anyway? I
understand passing in a pointer encoded error to the response handlers, but
the error routines take no action on a valid pointer and should never be
called that way.
Signed-off-by: Chris Leech <christopher.leech@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
In case of sequence offload, in fc_fcp_send_data(), the skb_fill_page_info()
called may end up adding more frags to the skb_shinfo(fp_skb(fp))->frags[],
exceeding SKB_MAX_FRAGS, this eventually corrupts the memory. I am adding the
FR_FRAME_SG_LEN back, but as SKB_MAX_FRAGS -1, leaving 1 for our fcoe_eof_crc
page. And send will be broken into multiple large sends if the frame already
contains more frags than skb handle.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Reported-by: Alex Lyakas <alexl@mellanox.co.il>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adds missing exch release when RRQ is accepted by calling
fc_seq_ls_acc. Adds common exch release for fc_exch_els_rrq
by use of out label.
Reported-by: Alex Lyakas <alexl@mellanox.co.il>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Initializing these libfc globals per lport could mess up exch
allocation/free for existing lport.
So this patch moves their initialization to fc_setup_exch_mgr
so that these globals gets initialized only once for libfc.
Reported-by: Alex Lyakas <alexl@mellanox.co.il>
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
It's possible and harmless to get FLOGI timeouts
while in RESET state. Don't do a WARN_ON in that case.
Also, split out the other WARN_ONs in fc_lport_timeout, so
we can tell which one is hit by its line number.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Fix minor errors.
A debug message said an RLIR was received instead of ECHO.
"Expected" was misspelled in several places.
Fix a type cast from u32 to __be32.
Rob, Some of these may have been also taken care of in your
other doc cleanup patch. Feel free to fold them in.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This bug is exposed when there is a link flap in LLD. Particularly, when it
happens right after a SCSI write command is sent out, no FCP_DATA is sent,
causing fsp->status_code to be set as FC_DATA_UNDRUN in fc_fcp_complete_locked
even no SCSI status is received. Consequently, fc_io_compl treats this as DID_OK.
This results in SCSI returning successful to the initial I/O request even
there is no DATA actually sent. Particularly, if you run an I/O tool w/ data
verification on, the read back for verification is gonna fail.
This is fixed here by checking when FC_DATA_UNDRUN happens, SCSI status is
received w/ FC_SRB_RCV_STATUS set in fsp->state.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This argument isn't used, let's not pass it into the routine.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
These are a few functions that were not used by other
modules. They did not need to be exported so this patch
removes the EXPORT_SYMBOLS call for each.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
xid 0 was used as an indication of invalid xid before but now xid 0
can be used as a valid exchange i. This patch fixes the ddp completion
in fcp layer, i.e., in fc_fcp.c:fc_fcp_ddp_done() function, to make sure it
does not use xid 0 for indication of an invalid xid, instead, it now
uses use FC_XID_UNKNOWN for such indication.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
A received Fibre Channel ELS PRLI request contains a bit that
indicates whether the remote port supports certain retry processing
sequences. The test for this bit was somehow coded to use multiply
instead of AND!
This case would apply only for target mode operation, and it is
unlikely to be noticed as an initiator.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
I saw an lport debug message from the exchange manager saying:
"lport 70500: Received response for out of range oxid:ffff"
A trace showed this was a BA_RJT sent due to an incoming ABTS
which arrived on an unknown exchange. So, the sender of the
BA_RJT was in error, but in this case, both the initiator and
responder were the same machine.
The OX_ID and RX_ID should not have been reversed in this case.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When an RSCN indicates changes to individual remote ports,
don't blindly log them out and then back in. Instead, determine
whether they're still in the directory, by doing GPN_ID.
If that is successful, call login, which will send ADISC and reverify,
otherwise, call logoff. Perhaps we should just delete the rport,
not send LOGO, but it seems safer.
Also, fix a possible issue where if a mix of records in the RSCN
cause us to queue disc_ports for disc_single and then we decide
to do full rediscovery, we leak memory for those disc_ports queued.
So, go through the list of disc_ports even if doing full discovery.
Free the disc_ports in any case. If any of the disc_single() calls
return error, do a full discovery.
The ability to fill in GPN_ID requests was added to fc_ct_fill().
For this, it needs the FC_ID to be passed in as an arg.
The did parameter for fc_elsct_send() is used for that, since the
actual D_DID will always be 0xfffffc for all CT requests so far.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The local port facility has been replying to ADISC requests without
looking to see if the remote port is logged in. This is incorrect.
An ADISC request requires PLOGI first. It should be rejected if
the sending remote port is not logged in.
This is like other incoming requests that require login, all of
which should be handled in the remote port module.
Move the ADISC request handling from fc_lport.c to fc_rport.c.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When rport_login is called on an rport that is already thought
to be logged in, use ADISC. If that fails, redo PLOGI.
This is less disruptive after fabric changes that don't affect
the state of the target.
Implement the sending of ADISC via fc_els_fill.
Add ADISC state to the rport state machine. This is entered from READY
and returns to READY after successful completion. If it fails, the rport
is either logged off and deleted or re-does PLOGI.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
fc_rport_logo_resp() had a call to fc_rport_enter_rtv() if the
LOGO was accepted. This must've been a copy/paste mistake, but
it didn't matter since we don't stay in the LOGO state long enough
to hit this code.
Change fc_rport_logo_resp() to just enter the delete state
no matter what.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
After a quick link flap, a target was seen to send us a LOGO.
Apparently, it saw an RSCN reporting that we had dropped out of the
fabric after we had logged back into it.
This is likely in larger fabrics (more than 2 FC switches) after
a quick link flap at the initiator. Each link transition causes
an port-specific RSCN to the target. After the link comes back up,
the initiator successfully discovers and does a PLOGI to the target
before the target sees the first RSCN reporting the initiator is gone,
and it sends a LOGO. The target may see a subsequent RSCN saying the
port is back, but probably wouldn't send a PLOGI and leaves it
up to the initiator to re-login.
An RSCN can be delayed by the switches due to software layers but a
PLOGI is forwarded in hardware causing the PLOGI to beat the RSCN.
If a remote port is in the discovered set and sends a LOGO, re-login to it.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When receiving an ELS request, if the request isn't recognized,
the unsupported operation error should be given even if the port
is not found or not logged in.
Also, the LOGO request shouldn't give the login-required explanation.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
libfc receives PLOGIs from switches which are trying to discover what
kind of devices are present, and from other initiators to find out
if we're a target.
As an initiator, some argue we don't need to handle incoming PLOGI
requests, and we currently reject them from unknown remote ports,
but accept them is we're in the middle of a PLOGI to the remote port.
For eventual target implementations, we want to handle them always.
For incoming PLOGI, don't fail if the rport_priv doesn't exist.
Just create it and go become READY without going through PRLI. If
PRLI occurs, then our roles will be set and we'll become READY again.
Also, allow incoming PRLI in RTV state.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Improve lport and rport debug messages to indicate whether
the response is LS_ACC, LS_RJT, closed, or timeout.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The rport_lookup function must be called while holding the disc_mutex.
Otherwise, the rdata could be deleted just after that by another thread.
All callers now check the state after grabbing the rdata rp_mutex.
Even though rport_lookup skips ports in DELETE state, it does that
without holding the rdata rp_mutex, so that the state may change.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This moves the remote port lookup for incoming ELS requests into
fc_rport.c, in preparation for handing PLOGI and LOGO from
unknown rports.
This changes the arg to rport_recv_req from an rdata to an lport.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Don't trust previous roles, reset them when we receive a PRLI.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Currently these values are initialized by the callers. This was exposed
by a later patch that adds PLOGI request support. The patch failed to
initialize the new remote port's roles and it caused problems. This patch
has the rport_create routine initialize the identifiers and then the
callers can override them with real values.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
It's possible to "restart" discovery before it was started if
an RSCN is received early enough. We were jumping to 0
due to the disc_callback function pointer not getting set.
Don't restart discovery if disc_callback is NULL.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The discovery code had a special-case for the point-to-point mode,
which used a bunch of code that wasn't really needed.
Now that rport_create adds the rport to the discovery list,
completely skip discovery for the point-to-point case.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
In fc_disc_gpn_ft_parse(), after fc_disc_done() is called, the
disc state is changed by setting buf_len = 0. This is wrong
since the discovery may have restarted. Instead, return
after calling fc_disc_done.
Also, return an error on memory allocation failure.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Currently fc_disc_timeout() restarts discovery only if it is not pending.
When the timer is scheduled, the discovery is left pending, so the
timeout never restarts it.
Fix by not checking for pending in the timeout handler.
If discovery is stopped and restarted in the meantime, the timeout will
be canceled.
Also, when a new discovery is started, the retry count wasn't cleared.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
On some switches, an empty zone causes GPN_FT to be rejected
with reason 9 (unable) explanation 7 (FC-4 types not registered),
which causes discovery to be retried endlessly. Treat this as
just an empty response and consider discovery complete.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Abhijeet Joglekar wrote: "In gpn_ft_resp, if the payload is short,
or unexpected response or out of sequence frame, then we just
return and do nothing. We should either enter fc_disc_done()
with DISC_EV_FAIL which will then restart any queued discovery
requests or call lport module which will reset local port,
or we should call fc_disc_error() so that the gpn_ft is retried.
The situation as is causes discovery to remain pending and never
get restarted, in these rare cases. We saw this due to a coding
bug in fc_disc before. The only ways it could happen would be
bugs, packet corruption or an FC fabric problem.
Change it to fail discovery. The local port will restart
discovery, although it probably should just give up until
the next link flap.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Code cleanup for fc_disc_gpn_ft_resp().
Some of the fc_disc.c code was poorly formatted. For example, some lines
in fc_disc.c were unnecessarily truncated and the buf variable could
be eliminated.
Also moved the increment of seq_count into fc_disc_gpn_ft_parse(), to
avoid doing it separately before each call.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When an RSCN is received during fabric discovery, it restarts.
After the restart, disc->seq_count was incremented, so when
the first frame was received, it was considered "out of sequence".
That left the state disc->active, preventing further discoveries.
Change to advance the sequence count before parsing, so that it
won't be changed after a potential restart.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When receiving an RSCN, do not log off all rports. This is
extremely disruptive. If, after the GPN_FT response, some
rports haven't been listed, delete them.
Add field disc_id to structs fc_rport_priv and fc_disc.
disc_id is an arbitrary serial number used to identify the
rports found by the latest discovery. This eliminates the need
to go through the rport list when restarting discovery.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Move the libfc remote port lookup function into fc_rport.c.
This seems like the best place for it.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Since the rport list maintenance is now done in the rport module,
the callback (and ops) are usually not necessary.
Allow rdata->ops to be left NULL if nothing needs
to be done in an event callback.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
For future discovery patches, change rport_create to return a previously
created rport_priv that has the FC_ID as long as it isn't in deleted state.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The list of remote ports (struct fc_rport_priv) has been
maintained by the discovery module. In preparation for having
lport->tt.rport_create() do a lookup first, maintain the
rports list in the rport module. It will still be protected
by the disc_mutex.
The DNS rport is an exception for until after further patches.
For now, do not add it to the list.
The point-to-point rport will be in the discovery list.
So at shutdown, it doesn't need to be separately logged out.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The lport rport callback can only be called for the dNS rport,
since its the only rport who's ops point to that function.
Remove unnecessary checking and debug messages.
Put the locking outside the switch statement as a simplification.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Don't print large negative decimal numbers for frame pointers in
the debug messages from fc_rport_error(). Just print 0 if its a
frame pointer, and print the error numbers as positive.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Delete unused disc->delay element.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
There was no need to have the discovery status stored in struct fc_disc.
Change fc_disc_done() to take the discovery status as an argument
and just pass it on to the discovery callback.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When a remote port becomes ready and a LOGO is received before
the READY event is in rport_work waiting on the mutex, the
event is changed to LOGO and the work queued, so both the
calls to rport_work see the LOGO event, and both try to do
the list_del(), causing a crash.
Don't change the event if it is already set.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Don't create a "dummy" remote port to go with fc_rport_priv.
Make the rport truly optional by allocating fc_rport_priv separately
and not requiring a dummy rport to be there if we haven't yet done
fc_remote_port_add().
The fc_rport_libfc_priv remains as a structure attached to the
rport for I/O purposes.
Be sure to hold references on rdata when the lock is dropped in
fc_rport_work().
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Remote ports will become READY more than once after
ADISC is implemented in a later patch.
The event callback that has been called "CREATED" will mean "READY".
Rename it now in preparation for those changes.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
This is a cleanup without semantic changes to use a switch
statement instead of a series of if-statements in fc_rport_work(),
and to move some declarations up to the top.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Allow a struct fc_rport_priv to have no fc_rport associated with it.
This sets up to remove the need for "rogue" rports.
Add a few fields to fc_rport_priv that are needed before the fc_rport
is created. These are the ids, maxframe_size, classes, and rport pointer.
Remove the macro PRIV_TO_RPORT(). Just use rdata->rport where appropriate.
To take the place of the get_device()/put_device ops that were used to
hold both the rport and rdata, add a reference count to rdata structures
using kref. When kref_get decrements the refcount to zero, a new template
function releasing the rdata should be called. This will take care of
freeing the rdata and releasing the hold on the rport (for now). After
subsequent patches make the rport truly optional, this release function
will simply free the rdata.
Remove the simple inline function fc_rport_set_name(), which becomes
semanticly ambiguous otherwise. The caller will set the port_name and
node_name in the rdata->Ids, which will later be copied to the rport
when it its created.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
tt.elsct_send is used by both FCP and by the rport state machine.
After further patches, these two modules will use different
structures for the remote port.
So, change elsct_send to use the FC_ID instead of the fc_rport_priv
as its argument. It currently only uses the FC_ID anyway.
For CT requests the destination FC_ID is still implicitly 0xfffffc.
After further patches the did arg on CT requests will be used to
specify the FC_ID being inquired about for GPN_ID or other queries.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The rport and discovery modules deal with remote ports
before fc_remote_port_add() can be done, because the
full set of rport identifiers is not known at early stages.
In preparation for splitting the fc_rport/fc_rport_priv allocation,
make fc_rport_priv the primary interface for the remote port and
discovery engines.
The FCP / SCSI layers still deal with fc_rport and
fc_rport_libfc_priv, however.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The interface for lport->tt.rport_create() takes a fc_disc_port arg,
which is unnatural for most calls. The only reason for this was
to avoid passing in the local port as an argument, but otherwise
added to complexity.
Simplify by just using lport and fc_rport_identifiers.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
While the I/O and LLD interfaces use fc_rport_libfc_priv, the
disc and rport interfaces will use fc_rport_priv, which will
be separately allocated.
Change the disc and rport usage of fc_rport_libfc_priv to fc_rport_priv.
Use #define temporarily to make both names equivalent until a
subsequent patch splits them.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
1. Updates fcoe_rcv() to queue incoming frames to the fcoe per
cpu thread on which this frame's exch was originated and simply
use current cpu for request exch not originated by initiator.
It is redundant to add this code under CONFIG_SMP, so removes
CONFIG_SMP uses around this code.
2. Updates fc_exch_em_alloc, fc_exch_delete, fc_exch_find to use
per cpu exch pools, here fc_exch_delete is rename of older
fc_exch_mgr_delete_ep since ep/exch are now deleted in pools
of EM and so brief new name is sufficient and better name.
Updates these functions to map exch id to their index into exch
pool using fc_cpu_mask, fc_cpu_order and EM min_xid.
This mapping is as per detailed explanation about this in
last patch and basically this is just as lower fc_cpu_mask
bits of exch id as cpu number and upper bit sum of EM min_xid
and exch index in pool.
Uses pool next_index to keep track of exch allocation from
pool along with pool_max_index as upper bound of exches array
in pool.
3. Adds exch pool ptr to fc_exch to free exch to its pool in
fc_exch_delete.
4. Updates fc_exch_mgr_reset to reset all exch pools of an EM,
this required adding fc_exch_pool_reset func to reset exches
in pool and then have fc_exch_mgr_reset call fc_exch_pool_reset
for each pool within each EM for a lport.
5. Removes no longer needed exches array, em_lock, next_xid, and
total_exches from struct fc_exch_mgr, these are not needed after
use of per cpu exch pool, also removes not used max_read,
last_read from struct fc_exch_mgr.
6. Updates locking notes for exch pool lock with fc_exch lock and
uses pool lock in exch allocation, lookup and reset.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adds per cpu exch pool for these reasons:-
1. Currently an EM instance is shared across all cpus to manage
all exches for all cpus. This required em_lock across all
cpus for an exch alloc, free, lookup and reset each frame
and that made em_lock expensive, so instead having per cpu
exch pool with their own per cpu pool lock will likely reduce
locking contention in fast path for an exch alloc, free and
lookup.
2. Per cpu exch pool will likely improve cache hit ratio since
all frames of an exch will be processed on the same cpu on
which exch originated.
This patch is only prep work to help in keeping complexity of next
patch low, so this patch only sets up per cpu exch pool and related
helper funcs to be used by next patch. The next patch fully makes
use of per cpu exch pool in all code paths ie. tx, rx and reset.
Divides per EM exch id range equally across all cpus to setup per
cpu exch pool. This division is such that lower bits of exch id
carries cpu number info on which exch originated, later a simple
bitwise AND operation on exch id of incoming frame with fc_cpu_mask
retrieves cpu number info to direct all frames to same cpu on which
exch originated. This required a global fc_cpu_mask and fc_cpu_order
initialized to max possible cpus number nr_cpu_ids rounded up to 2's
power, this will be used in mapping exch id and exch ptr array
index in pool during exch allocation, find or reset code paths.
Adds a check in fc_exch_mgr_alloc() to ensure specified min_xid
lower bits are zero since these bits are used to carry cpu info.
Adds and initializes struct fc_exch_pool with all required fields
to manage exches in pool.
Allocates per cpu struct fc_exch_pool with memory for exches array
for range of exches per pool. The exches array memory is followed
by struct fc_exch_pool.
Adds fc_exch_ptr_get/set() helper functions to get/set exch ptr in
pool exches array at specified array index.
Increases default FCOE_MAX_XID to 0x0FFF from 0x07EF, so that more
exches are available per cpu after above described exch id range
division across all cpus to each pool.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The statement reads, "Exchange timed out, notifying the upper layer",
however, this statement is printed whenever the timer is armed. This
is confusing to someone debugging the code because every time an
exchange is initialized, there is an incorrect statement stating that
the timer has already timed out. This patch changes the statement to
read, "Exchange timer armed" which is more accurate.
This patch also adds a debug statement in the timeout handler to
properly indicate that the exchange has timed out.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
I don't believe this check is needed any more in the current kernel, which,
if I understand correctly, is for compound page where only the first page
is supposed to get ref-counted.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
FC_FRAME_SG_LEN is 4 which is too small when offload is enabled. Actually, the
WARN_ON() in fc_fcp_send_data() should be:
WARN_ON(skb_shinfo(fp_skb(fp))->nr_frags > MAX_SKB_FRAGS);
But since we will not get anything more than 64K anyway, so there is no need
to do this anyway here. Therefore, I am getting rid of FC_FRAME_SG_LEN here
and the WARN_ON here.
Signed-off-by: Yi Zou <yi.zou@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Updates fcoe_em_config to allocate a single instance of sharable offload
EM for supported lp->lro_xid per eth device, and then share this EM
for subsequently more lports creation on same eth device (e.g when using
VLAN).
Adds tiny fcoe_oem_match function for offload EM to return true for read
types IO to have read IO exchanges allocated from offload shared EM.
Removes fc_em_alloc_xid function completely which was needed to manage
two xid ranges within a EM, this is not needed any more with allocation
of separate sharable offload EM per eth device. Instead this patch adds
simple xid allocation logic to manage single xid range.
Adds fc_exch_em_alloc with mp->next_xid as cursor to allocate new xid
from single xid range of EM, uses mp->next_xid instead removed mp->last_xid
which slightly increase probability of finding empty xid on exch allocation.
Removes restriction of not allowing use of xid zero along with changing
two xid range change to single xid range.
Makes fc_fcp_ddp_setup calling conditional to only xid allocated from
shared offload EM.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Modifies current code to use EM anchor list in EM allocation, EM free,
EM reset, exch allocation and exch lookup code paths.
1. Modifies fc_exch_mgr_alloc to accept EM match function and then
have allocated EM added to the lport using fc_exch_mgr_add API
while also updating EM kref for newly added EM.
2. Updates fc_exch_mgr_free API to accept only lport pointer instead
EM and then have this API free all EMs of the lport from EM anchor
list.
3. Removes single lport pointer link from the EM, which was used in
associating lport pointer in newly allocated exchange. Instead have
lport pointer passed along new exchange allocation call path and
then store passed lport pointer in newly allocated exchange, this
will allow a single EM instance to be used across more than one
lport and used in EM reset to reset only lport specific exchanges.
4. Modifies fc_exch_mgr_reset to reset all EMs from the EM anchor list
of the lport, adds additional exch lport pointer (ep->lp) check for
shared EM case to reset exchange specific to a lport requested reset.
5. Updates exch allocation API fc_exch_alloc to use EM anchor list and
its anchor match func pointer. The fc_exch_alloc will walk the list
of EMs until it finds a match, a match will be either null match
func pointer or call to match function returning true value.
6. Updates fc_exch_recv to accept incoming frame on local port using
only lport pointer and frame pointer without specifying EM instance
of incoming frame. Instead modified fc_exch_recv to locate EM for the
incoming frame by matching xid of incoming frame against a EM xid range.
This change was required to use EM list in libfc Rx path and after this
change the lport fc_exch_mgr pointer emp is not needed anymore, so
removed emp pointer.
7. Updates fnic for removed lport emp pointer and above modified libfc APIs
fc_exch_recv, fc_exch_mgr_alloc and fc_exch_mgr_free.
8. Removes exch_get and exch_put from libfc_function_template as these
are no longer needed with EM anchor list and its match function use.
Also removes its default function fc_exch_get.
A defect this patch introduced regarding the libfc initialization order in
the fnic driver was fixed by Joe Eykholt <jeykholt@cisco.com>.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Currently there is a 1:1 relationship between the lport
and exchange manager. This macro takes an EM as an argument
and determines the lport from it. However, later patches
will use an EM list per lport, so we will no longer have
this 1:1 relationship- this macro must change.
The FC_EM_DBG macro is rarely used. There are four callers,
two can use FC_LPORT_DBG instead and two can be removed
since they're not necessary. This patch makes those changes
and removes the macro.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Adds EM list using a anchor struct fc_exch_mgr_anchor, anchor is used
to allow same EM instance sharing across more than one lport on a eth
device, this implementation is per discussed design posted at
http://www.open-fcoe.org/pipermail/devel/2009-June/002566.html.
The shared EM is required for multiple lports on eth device when
using multiple VLANs or NPIV.
Adds fc_exch_mgr_add API to add a EM to the lport and fc_exch_mgr_del
API to delete previously added EM.
Also adds function fc_exch_mgr_destroy() to destroy allocated EM.
The kref is added to the EM to keep track of EM usage count, the EM is
destroyed when no longer in use upon kref reaching to zero.
The caller can specify match function to fc_exch_mgr_add, this
will be used in determining exchange allocation from its EM or not.
Moved calling of fcoe_em_config below fcoe_libfc_config calling,
so that list head lp->ema_list is initialized before configuring
EM.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The timer for rport retries wasn't getting canceled, and
would occasionally go off after the module was unloaded.
Add logic to cancel the timer in fc_rport_work().
Since we cancel the timer before deleting the rdata,
it is no longer necessary to do a get_device() for the pending timer.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
fc_rport_logoff drops the rport lock in order to cancel work
that may be pending. This is undesirable as the state can
completely change, and the caller may not expect that the
lock could've been dropped.
If there is work pending, it will acquire the rdata mutex and
so we're protected and can change the event from READY to DELETE.
Queue the work only if there is no event already pending.
There were a couple other cases where the state was set to
DELETE and work queued, even though the state may have already
been DELETE. Fix these using a common function fc_rport_enter_delete().
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
State RPORT_ST_NONE was intented to be an invalid state (0), never used.
This was a misguided attempt to be sure it was always initialized.
Having an extra state meaning nothing requires switch statements to
have a case covering that state.
State NONE has been used instead to mean the remote port is being deleted.
Changing the name to RPORT_ST_DELETE.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
We saw periodic messages like:
WARNING: at drivers/scsi/libfc/fc_exch.c:825 fc_seq_start_next+0x30/0x4b
This was due to trying to allocate a sequence in a request handler
when the exchange had been reset.
Delete the WARN_ON.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
During an fcoe module unload, we saw a problem where fc_rport_work()
finds the lport has been freed. The rdata points to an area
containing 0x6b6b6b6b... the pool poison value from kmem_free().
In fcoe_if_destroy() we call fc_fabric_logoff() then fc_lport_destroy().
fc_fabric_logoff() flushes the remote port work, but we're still receiving
requests, and an RSCN or PLOGI arrives which creates more rports.
Note that although the LLD also checks link_up, it doesn't do it
under the lport mutex, so it can deliver frames to
fc_lport_recv_req() even after link_up is cleared.
So, re-check link_up there.
We need to flush the rports by calling disc_stop_final()
after we clear link_up.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
When removing the fcoe module, several lports were being shut down
through fc_lport_fabric_logoff().
Occasionally, one would enter reset state before fc_lport_destroy()
was called, and since link_up was still true, it would log back in.
If we just clear link_up earlier, then we wouldn't be accepting LOGO
requests from other initiators while we are shutting down.
Fix by changing the LOGO response handler to enter DISABLED instead
of RESET. Add an fc_lport_enter_disabled() function which does
what fc_lport_enter_reset() did, except it doesn't proceed to FLOGI state.
Move the code that was common between fc_lport_enter_reset() and
fc_lport_enter_disabled() into a new fc_lport_reset_locked() function.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
The state NONE was meant to be invalid, but has been used as
the initial state. Rename it to be DISABLED, as more descriptive.
Further patches will make it the like the RESET state, except
it won't transition to FLOGI until fc_lport_fabric_login() is called.
Signed-off-by: Joe Eykholt <jeykholt@cisco.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
Currently the fc_exch_rrq is called with fc_exch's ex_lock held.
The fc_exch_rrq allocates new exch and that requires taking
ex_lock again after EM lock. This locking order causes warning,
see more details on this warning at :-
http://www.open-fcoe.org/pipermail/devel/2009-July/003251.html
This patch fixes this by dropping the ex_lock before calling
fc_exch_rrq().
The fc_exch_rrq needs to grab ex_lock lock again to schedule
RRQ retry and in the meanwhile fc_exch_reset could occur before
ex_lock is grabbed inside fc_exch_rrq. So to handle this case,
this patch adds additional check to detect fc_exch_reset after
ex_lock acquired and in case the fc_exch_reset occurred then
abandons the RRQ retry and releases the exch.
Signed-off-by: Vasu Dev <vasu.dev@intel.com>
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
This patch adds the /sys/module/libfc/parameters/debug_logging
file to sysfs as a module parameter. It accepts an integer
bitmask for logging. Currently it supports:
bit
LSB 0 = general libfc debugging
1 = lport debugging
2 = disc debugging
3 = rport debugging
4 = fcp debugging
5 = EM debugging
6 = exch/seq debugging
7 = scsi logging (mostly error handling)
the other bits are not used at this time.
The patch converts all of the libfc source files to use
these new macros and removes the old FC_DBG macro.
Signed-off-by: Robert Love <robert.w.love@intel.com>
Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
Joe Eykholt