Previously the driver and userspace code handled the case of 1 subport
somewhat inconsistently.  The new interpretation of this situation is
that if one subport is requested, the driver turns on the subport
mechanism and arranges for the port to be "shared" by one process.  In
normal use the userspace library does not use this configuration and
instead arranges for the port not to be shared at all.  This
particular idiom can be useful for testing purposes.
Signed-off-by: NMark Debbage <mark.debbage@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

When subports are required to run a program, this patch checks that
the driver and the userspace library have compatible subport
implementations.  This is achieved through checks on the swminor
version field built into the driver and userspace library.  Bad
combinations are reported through syslog and result in an error when
opening the port.
Signed-off-by: NMark Debbage <mark.debbage@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

A duplicate RDMA read request can fool the responder into NAKing a new
RDMA read request because the responder wasn't keeping track of
whether the queue of RDMA read requests had been sent at least once.
For example, requester sends 4 2K byte RDMA read requests, times out,
and resends the first, then sees the 4 responses, then sends a 5th
RDMA read or atomic operation.  The responder sees the 4 requests,
sends 4 responses, sees the resent 1st request, rewinds the queue,
then sees the 5th request but thinks the queue is full and that the
requester is invalidly sending a 5th new request.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The code to copy data from the receive queue buffers to the IB SGEs
doesn't check the SGE length, only the memory region/page length when
copying data.  This could overwrite parts of the user's memory that
were not intended to be written.  It can only happen if multiple SGEs
are used to describe a receive buffer which almost never happens in
practice.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The send function is called when posting new send work requests.
There is no point in trying to send a packet if the QP is already
waiting for a HW send buffer so don't clear the busy bit until the
buffer available interrupt happens.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

A RDMA read response or atomic response can ACK earlier sends and RDMA
writes.  In this case, the wrong work request pointer was being used
to store the read first response or atomic result.  Also, if a RDMA
read request is retried, the code to compute which request to resend
was incorrect.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

This centralizes the use of the abort functionality, removes the
unneeded buffer cancel (abort does the same thing), sets up to ignore
launch errors after abort, same as cancel.  We need abort on exit from
freeze mode to avoid having buffers stuck in the busy state, if a user
process happened to complete the send while we were in freeze mode
doing the recovery.
Signed-off-by: NDave Olson <dave.olson@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The values passed have never been right for iba 6120 chips, but just
happened to work.  We needed to select the right buffer offset in the
chip (both are in same register), and the total length was wrong also,
but was covered by the rounding up.
Signed-off-by: NDave Olson <dave.olson@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Define pkt rcvd 'type' in a way consistent with HW spec and chips.

The hardware considers received packets of type 0 to be expected, and
type 1 to be eager. The driver was calling the ipath_f_put_tid
functions using a variable called 'type' set to 0 for eager and to 1
for expected packets.  Worse, the iba6110 and iba6120 drivers used
those values inconsistently.  This was quite confusing.  Now
everything is consistent with the hardware.
Signed-off-by: NDave Olson <dave.olson@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

According to chapter 17.2.8.1.1, QPs start in the migrated state and
should send packets with the M bit set in the BTH.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

This patch fixes a minor bug where the wrong QP was checked for a send
work request that should wait for an RNR timeout.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

This patch fixes a bug introduced when moving some code around for
readability.

Setting the wqe pointer at the end of the function is a NOP since it
isn't used.  Move it back to where it is used.
Signed-off-by: NRalph Campbell <ralph.campbell@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

In ipath_query_device(), some of the struct ib_device_attr fields were
not being initialized.
Signed-off-by: NRobert Walsh <robert.walsh@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Although our chip supports 4K MTUs, our driver doesn't yet support
this feature, so limit the maximum MTU to 2K until we get support for
4K MTUs implemented.
Signed-off-by: NRobert Walsh <robert.walsh@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Recognize IBA 6110 Revision 4: same feature set, etc. as earlier revisions.
Signed-off-by: NDave Olson <dave.olson@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

We currently track various errors, now we enhance that capability by
logging some of them to EEPROM.  We also now log a cumulative "active"
time defined by traffic though the InfiniPath HCA beyond the normal SM
traffic.
Signed-off-by: NMichael Albaugh <michael.albaugh@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The IPATH_RUNTIME_PBC_REWRITE and the IPATH_RUNTIME_LOOSE_DMA_ALIGN
flags were not ever implemented correctly and did not turn out to be
necessary.  Remove the last vestiges of these flags but mark the spot
with a comment to remind us to not reuse these flags in the interest
of binary compatibility.  The INFINIPATH_XGXS_SUPPRESS_ARMLAUNCH_ERR
bit was also not found to be useful, so it was dropped in the cleanup
as well.
Signed-off-by: NJohn Gregor <john.gregor@qlogic.com>
Signed-off-by: NArthur Jones <arthur.jones@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The new LED blinking interface adds more contention for the
unprotected GPIO pins that were already shared, though not commonly at
the same time.  We add locks to the accesses to these pins so that
Read-Modify-Write is now safe.  Some of these locks are added at
interrupt context, so we shadow the registers which drive and inspect
these pins to avoid the mmio read/writes.  This mitigates the effects
of the locks and hastens us through the interrupt.

Add locking and always use shadows for registers controlling GPIO pins
(ExtCtrl and GPIOout). The use of shadows implies doing less I/O,
which can make I2C operation too fast on some platforms. An explicit
udelay(1) in SCL manipulation fixes that.
Signed-off-by: NMichael Albaugh <michael.albaugh@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

When we want to find an InfiniPath HCA in a rack of nodes, it is often
expeditious to blink the status LEDs via a userspace /sys file.

A write-only led_override "file" is published per device. Writes to
this file are interpreted as (string form) numbers, and the resulting
value sent to ipath_set_led_override(). The upper eight bits are
interpretted as a 4.4 fixed-point "frequency in Hertz", and the bottom
two 4-bit values are alternately (D0..3, then D4..7) used by the
board-specific LED-setting function to override the normal state.
Signed-off-by: NMichael Albaugh <michael.albaugh@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Signed-off-by: NBryan O'Sullivan <bryan.osullivan@qlogic.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

mlx4_ib.h uses struct mutex, so although <linux/mutex.h> seems to be
pulled in indirectly by one of the headers it includes, the right
thing is to include <linux/mutex.h> directly.
Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Refactor the ehca changes from commit ed23a727 ("IB: Return "maybe
missed event" hint from ib_req_notify_cq()") so the queue arithmetic
is done in slightly fewer lines.  Also, move the spinlock flags into
the block they're used in.
Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

We need to keep a spare entry in the SRQ so that there always is a
next WQE available when posting receives (so that we can tell the
difference between a full queue and an empty queue).  So subtract 1
from the value HW gives us before reporting the limit on SRQ entries
to consumers.

Found by Mellanox QA.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Inline data segments in send WQEs are not allowed to cross a 64 byte
boundary.  We use inline data segments to hold the UD headers for MLX
QPs (QP0 and QP1).  A send with GRH on QP1 will have a UD header that
is too big to fit in a single inline data segment without crossing a
64 byte boundary, so split the header into two inline data segments.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Upcoming firmware introduces command interface revision 3, which
changes the way port capabilities are queried and set.  Update the
driver to handle both the new and old command interfaces by adding a
new MLX4_FLAG_OLD_PORT_CMDS that it is set after querying the firmware
interface revision and then using the correct interface based on the
setting of the flag.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

When compacting CQ entries, we need to set the correct value of the
ownership bit in case the value is different between the index we copy
the CQE from and the index we copy it to.

Found by Ronni Zimmerman of Mellanox.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The calculation of max_inline_data in set_kernel_sq_size() is bogus,
since it doesn't take into account the fact that inline segments may
not cross a 64-byte boundary, and hence multiple inline segments will
probably need to be used to post large inline sends.

We don't support inline sends for kernel QPs anyway, so there's no
point in doing this calculation anyway, since the field is just zeroed
out a little later.  So just delete the bogus calculation.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

New ConnectX firmware introduces FW command interface revision 2,
which requires that for each QP, a chunk of send queue entries (the
"headroom") is kept marked as invalid, so that the HCA doesn't get
confused if it prefetches entries that haven't been posted yet.  Add
code to the driver to do this, and also update the user ABI so that
userspace can request that the prefetcher be turned off for userspace
QPs (we just leave the prefetcher on for all kernel QPs).

Unfortunately, marking send queue entries this way is confuses older
firmware, so we change the driver to allow only FW command interface
revisions 2.  This means that users will have to update their firmware
to work with the new driver, but the firmware is changing quickly and
the old firmware has lots of other bugs anyway, so this shouldn't be too
big a deal.

Based on a patch from Jack Morgenstein <jackm@dev.mellanox.co.il>.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Doing max(1, foo) where foo is u32 generates a warning, because 1 is a
signed constant.  Fix this by using 1U instead.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Cast the increment added to wq->tail when send completions are
processed to u16 to avoid using wrong values caused by standard
integer promotions.

The same bug was fixed in libmlx4 by Eli Cohen <eli@mellanox.co.il>.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

QPs attached to an SRQ must never have their own RQ, and QPs not
attached to SRQs must have an RQ with at least 1 entry.  Enforce all
of this in set_rq_size().

Based on a patch by Eli Cohen <eli@mellanox.co.il>.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

The code in __mlx4_ib_modify_qp() overwrites context->params1 after
the RNR retry parameter is ORed in, which results in the RNR retry
parameter always being set to 0.  Fix this by moving where we OR in
the value to later in the function, after the initial assignment of
context->params1.

Found by the Mellanox firmware group.
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

This patch converts the ipv4_devconf config members (everything except
sysctl) to an array.  This allows easier manipulation which will be
needed later on to provide better management of default config values.
Signed-off-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

s/signifant/significant/
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

mthca_free_err_wqe() currently treats both send and receive CQEs
identically if a QP is using an SRQ.  But for Tavor hardware, send
CQEs with error can be chained together even if the RQ is part of SRQ,
so we may miss some CQEs.

Fix by following the WQE chain for all send CQEs even for non-SRQ QPs.

This fixes crashes in IPoIB CM:
<https://bugs.openfabrics.org//show_bug.cgi?id=604>
Signed-off-by: NMichael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

Due to a typo, the driver was reporting the wrong number of "actual send
WRs" after ehca_create_qp().
Signed-off-by: NJoachim Fenkes <fenkes@de.ibm.com>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

We need to initialize the owner bit of send queue WQEs to hardware 
ownership whenever the QP is modified from reset to init, not just 
when the QP is first allocated.  This avoids having the hardware 
process stale WQEs when the QP is moved to reset but not destroyed and 
then modified to init again. 
Signed-off-by: NEli Cohen <eli@mellanox.co.il>
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

If a QP is attached to a shared receive queue (SRQ), then it doesn't
have a receive queue (RQ).  So don't allocate an RQ doorbell (or map a
doorbell from userspace for userspace QPs) for that QP.
Signed-off-by: NRoland Dreier <rolandd@cisco.com>

First thing mm.h does is including sched.h solely for can_do_mlock() inline
function which has "current" dereference inside. By dealing with can_do_mlock()
mm.h can be detached from sched.h which is good. See below, why.

This patch
a) removes unconditional inclusion of sched.h from mm.h
b) makes can_do_mlock() normal function in mm/mlock.c
c) exports can_do_mlock() to not break compilation
d) adds sched.h inclusions back to files that were getting it indirectly.
e) adds less bloated headers to some files (asm/signal.h, jiffies.h) that were
   getting them indirectly

Net result is:
a) mm.h users would get less code to open, read, preprocess, parse, ... if
   they don't need sched.h
b) sched.h stops being dependency for significant number of files:
   on x86_64 allmodconfig touching sched.h results in recompile of 4083 files,
   after patch it's only 3744 (-8.3%).

Cross-compile tested on

	all arm defconfigs, all mips defconfigs, all powerpc defconfigs,
	alpha alpha-up
	arm
	i386 i386-up i386-defconfig i386-allnoconfig
	ia64 ia64-up
	m68k
	mips
	parisc parisc-up
	powerpc powerpc-up
	s390 s390-up
	sparc sparc-up
	sparc64 sparc64-up
	um-x86_64
	x86_64 x86_64-up x86_64-defconfig x86_64-allnoconfig

as well as my two usual configs.
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Make mlx4_post_srq_recv() fail if the SRQ is full (head == tail).
Signed-off-by: NRoland Dreier <rolandd@cisco.com>