diff mbox

[v4,17/50] IB/hfi1: add PSM driver control/data path

Message ID 20150730191859.25256.28987.stgit@phlsvslse11.ph.intel.com (mailing list archive)
State Accepted
Headers show

Commit Message

Marciniszyn, Mike July 30, 2015, 7:18 p.m. UTC
Signed-off-by: Andrew Friedley <andrew.friedley@intel.com>
Signed-off-by: Arthur Kepner <arthur.kepner@intel.com>
Signed-off-by: Brendan Cunningham <brendan.cunningham@intel.com>
Signed-off-by: Brian Welty <brian.welty@intel.com>
Signed-off-by: Caz Yokoyama <caz.yokoyama@intel.com>
Signed-off-by: Dean Luick <dean.luick@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Easwar Hariharan <easwar.hariharan@intel.com>
Signed-off-by: Harish Chegondi <harish.chegondi@intel.com>
Signed-off-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Jim Snow <jim.m.snow@intel.com>
Signed-off-by: John Gregor <john.a.gregor@intel.com>
Signed-off-by: Jubin John <jubin.john@intel.com>
Signed-off-by: Kaike Wan <kaike.wan@intel.com>
Signed-off-by: Kevin Pine <kevin.pine@intel.com>
Signed-off-by: Kyle Liddell <kyle.liddell@intel.com>
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Signed-off-by: Mitko Haralanov <mitko.haralanov@intel.com>
Signed-off-by: Ravi Krishnaswamy <ravi.krishnaswamy@intel.com>
Signed-off-by: Sadanand Warrier <sadanand.warrier@intel.com>
Signed-off-by: Sanath Kumar <sanath.s.kumar@intel.com>
Signed-off-by: Sudeep Dutt <sudeep.dutt@intel.com>
Signed-off-by: Vlad Danushevsky <vladimir.danusevsky@intel.com>
---
 drivers/infiniband/hw/hfi1/file_ops.c | 2140 +++++++++++++++++++++++++++++++++
 1 file changed, 2140 insertions(+)
 create mode 100644 drivers/infiniband/hw/hfi1/file_ops.c


--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Comments

Jason Gunthorpe July 30, 2015, 7:40 p.m. UTC | #1
On Thu, Jul 30, 2015 at 03:18:59PM -0400, Mike Marciniszyn wrote:
> +static ssize_t hfi1_write_iter(struct kiocb *kiocb, struct iov_iter *from)
> +{
> +	struct hfi1_user_sdma_pkt_q *pq;
> +	struct hfi1_user_sdma_comp_q *cq;
> +	int ret = 0, done = 0, reqs = 0;
> +	unsigned long dim = from->nr_segs;

I thought you were getting rid of this?

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Marciniszyn, Mike July 30, 2015, 8:01 p.m. UTC | #2
> 
> I thought you were getting rid of this?
> 
> Jason

Doug wanted the v4 submitted as we currently have it.

Doug?

Mike
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Doug Ledford July 30, 2015, 9:42 p.m. UTC | #3
On 07/30/2015 04:01 PM, Marciniszyn, Mike wrote:
>>
>> I thought you were getting rid of this?
>>
>> Jason
> 
> Doug wanted the v4 submitted as we currently have it.

To be accurate, I said "If you want a chance at making 4.3, I need a
v4".  I didn't comment on whether or not any specific review comments
were addressed.

> Doug?

I have no problem with this code.  That Al finds the user space ABI for
this driver to be bizarre is neither here nor there to me.  Sure, this
file does not exhibit normal file API behavior.  Who cares?  It's not a
normal file in *any* sense of the word.  For example, the normal write
routine will never, ever accept just plain data.  It's always in the
form of a command.  If you don't have the right magic decoder ring, you
will get nothing but errors when trying to do something with this file.
 Much like /dev/infiniband/uverbs? files, it is a command interface, not
a raw data interface.  I actually think the fact that you guys use write
for a single command and writev/write_iter for a command queue is an
elegant solution to your particular needs.  The only reason Al threw a
hissy over it is because it tripped him up when he went to do the
conversion from writev to write_iter.  That's understandable.  So, some
clear documentation so someone like Al doesn't have to go reading
through sources and try to figure out what you are doing would be the
generally nice thing to do for other kernel generalists that might come
poking around this way.  Or, another option would be to drop the write
function altogether and just make all commands come through
writev/write_iter and if you only have one command, you only send one
element.  Regardless, those things can be cleaned up in follow on
patches, please do not resubmit this set for that.
Marciniszyn, Mike July 30, 2015, 10 p.m. UTC | #4
> On 07/30/2015 04:01 PM, Marciniszyn, Mike wrote:
> >>
> >> I thought you were getting rid of this?
> >>
> >> Jason
> >
> > Doug wanted the v4 submitted as we currently have it.
> 
> To be accurate, I said "If you want a chance at making 4.3, I need a v4".  I
> didn't comment on whether or not any specific review comments were
> addressed.
> 
> > Doug?
> 
> I have no problem with this code.  That Al finds the user space ABI for this
> driver to be bizarre is neither here nor there to me.  Sure, this file does not
> exhibit normal file API behavior.  Who cares?  It's not a normal file in *any*
> sense of the word.  For example, the normal write routine will never, ever
> accept just plain data.  It's always in the form of a command.  If you don't
> have the right magic decoder ring, you will get nothing but errors when
> trying to do something with this file.
>  Much like /dev/infiniband/uverbs? files, it is a command interface, not a
> raw data interface.  I actually think the fact that you guys use write for a
> single command and writev/write_iter for a command queue is an elegant
> solution to your particular needs.  The only reason Al threw a hissy over it is
> because it tripped him up when he went to do the conversion from writev to
> write_iter.  That's understandable.  So, some clear documentation so
> someone like Al doesn't have to go reading through sources and try to figure
> out what you are doing would be the generally nice thing to do for other
> kernel generalists that might come poking around this way.  Or, another
> option would be to drop the write function altogether and just make all
> commands come through writev/write_iter and if you only have one
> command, you only send one element.  Regardless, those things can be
> cleaned up in follow on patches, please do not resubmit this set for that.
> 

Jason,

I did ask you in http://marc.info/?l=linux-rdma&m=143707462806767&w=2 if you thought ioctl was ok.

Hearing nothing, we left the interface as it was.

I suspect (I lack the early history) that the ioctl BKL might have forced both uverbs and PSM to go this route.

Doug,

Where would be the appropriate location to document?  In the source itself?  Somewhere else?

Mike
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Doug Ledford July 30, 2015, 10:10 p.m. UTC | #5
On 07/30/2015 06:00 PM, Marciniszyn, Mike wrote:
>> On 07/30/2015 04:01 PM, Marciniszyn, Mike wrote:
>>>>
>>>> I thought you were getting rid of this?
>>>>
>>>> Jason
>>>
>>> Doug wanted the v4 submitted as we currently have it.
>>
>> To be accurate, I said "If you want a chance at making 4.3, I need a v4".  I
>> didn't comment on whether or not any specific review comments were
>> addressed.
>>
>>> Doug?
>>
>> I have no problem with this code.  That Al finds the user space ABI for this
>> driver to be bizarre is neither here nor there to me.  Sure, this file does not
>> exhibit normal file API behavior.  Who cares?  It's not a normal file in *any*
>> sense of the word.  For example, the normal write routine will never, ever
>> accept just plain data.  It's always in the form of a command.  If you don't
>> have the right magic decoder ring, you will get nothing but errors when
>> trying to do something with this file.
>>  Much like /dev/infiniband/uverbs? files, it is a command interface, not a
>> raw data interface.  I actually think the fact that you guys use write for a
>> single command and writev/write_iter for a command queue is an elegant
>> solution to your particular needs.  The only reason Al threw a hissy over it is
>> because it tripped him up when he went to do the conversion from writev to
>> write_iter.  That's understandable.  So, some clear documentation so
>> someone like Al doesn't have to go reading through sources and try to figure
>> out what you are doing would be the generally nice thing to do for other
>> kernel generalists that might come poking around this way.  Or, another
>> option would be to drop the write function altogether and just make all
>> commands come through writev/write_iter and if you only have one
>> command, you only send one element.  Regardless, those things can be
>> cleaned up in follow on patches, please do not resubmit this set for that.
>>
> 
> Jason,
> 
> I did ask you in http://marc.info/?l=linux-rdma&m=143707462806767&w=2 if you thought ioctl was ok.
> 
> Hearing nothing, we left the interface as it was.

I think the interface is fine as is, with the only thing I would do, if
*really* forced to by Al, would be to do as I suggested above and
convert all of the write cases to writev with a single element.

> I suspect (I lack the early history) that the ioctl BKL might have forced both uverbs and PSM to go this route.

An ioctl interface is not really designed for a queue of commands to be
sent in a single operation any more than any other interface is.  I
don't personally see a great benefit to it.

> 
> Doug,
> 
> Where would be the appropriate location to document?  In the source itself?  Somewhere else?

At the function for both write and writev (and a patch to update to
write_iter would be a good next step to keep us in line with qib),
document how each is used and point to the other one and point out that
they differ in their basic usage.  If Al had a clear comment saying "our
write function is used to pass a single command to our driver, and our
writev function is used to pass a queue of formatted commands, one per
element", he might have not written what he did in his commit message.
Jason Gunthorpe July 30, 2015, 10:30 p.m. UTC | #6
On Thu, Jul 30, 2015 at 06:10:41PM -0400, Doug Ledford wrote:
> > I did ask you in http://marc.info/?l=linux-rdma&m=143707462806767&w=2 if you thought ioctl was ok.
> > 
> > Hearing nothing, we left the interface as it was.

It isn't really fair to expect other people to design your UAPI for
you, Al didn't respond either. Nobody else but you understand what the
heck this is doing... I gave you lots of options and guidance. Pick
something that make sense.

> I think the interface is fine as is, with the only thing I would do, if
> *really* forced to by Al, would be to do as I suggested above and
> convert all of the write cases to writev with a single element.

The point of Al's comment was that write() and writev() do *different
things* and calling writev with a single iov entry is *NOT*
equivalent to write() with the same data.

That is fundamentally not how write and writev should behave - they
are supposed to be symmetrically interchangeable.

I don't think adding a comment makes the situation any better.

> they differ in their basic usage.  If Al had a clear comment saying "our
> write function is used to pass a single command to our driver, and our
> writev function is used to pass a queue of formatted commands,

Well, that would probably be fine, but that is not what is going on at
all....

Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Christoph Hellwig July 31, 2015, 7:31 a.m. UTC | #7
On Thu, Jul 30, 2015 at 05:42:16PM -0400, Doug Ledford wrote:
> I have no problem with this code.  That Al finds the user space ABI for
> this driver to be bizarre is neither here nor there to me.  Sure, this
> file does not exhibit normal file API behavior.  Who cares?

Everyone who cares about filesystem semantics does.

A NACK from me for a this features, and b) for trying to sneak it in
after a negative comment without cc to -fsdevel and Al.

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Al Viro July 31, 2015, 8:22 a.m. UTC | #8
On Fri, Jul 31, 2015 at 12:31:58AM -0700, Christoph Hellwig wrote:
> On Thu, Jul 30, 2015 at 05:42:16PM -0400, Doug Ledford wrote:
> > I have no problem with this code.  That Al finds the user space ABI for
> > this driver to be bizarre is neither here nor there to me.  Sure, this
> > file does not exhibit normal file API behavior.  Who cares?
> 
> Everyone who cares about filesystem semantics does.
> 
> A NACK from me for a this features, and b) for trying to sneak it in
> after a negative comment without cc to -fsdevel and Al.

FWIW, the lack of comments appears to have tripped Doug, judging by his
"another option would be to drop the write function altogether and just make
all commands come through writev/write_iter and if you only have one command,
you only send one element".

The thing is, ipath and qib (and AFAICS this one as well) have write(2) and
writev(2) take different and completely unrelated sets of commands.  On
the same file.  IOW, the effects of
	writev(fd, &(struct iovec){buf, len}, 1)
and
	write(fd, buf, len)
are not even remotely similar.  _That_ is the gratuitous weirdness I'd been
unhappy with.  And yes, it is gratuitous - it's trivial to have separate files
for separate command sets.

If you drop ->write() in there, you certainly lose the weirdness - along with
one of those command sets.  Sure, having individual iovecs correspond to
separate datagrams is fine; tons of drivers are like that.  qib and ipath
are unique, though, in having *two* command sets overloaded on the same file,
with write() vs. writev() acting as selector (BTW, single-element AIO
going like writev(), not like write()).

PS: I'm back after several weeks of being sick (recipe for fun: start with
40C in shade, get completely soaked in serious rain, then have half an hour
cab ride, with AC set to ~15C) and while I'd managed to get the mailbox
down to relatively sane size, I might have easily deleted more than I should
have.  If there had been anything important sent my way and I don't reply to 
it by Saturday, please resend.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Doug Ledford July 31, 2015, 2:19 p.m. UTC | #9
On 07/31/2015 04:22 AM, Al Viro wrote:
> On Fri, Jul 31, 2015 at 12:31:58AM -0700, Christoph Hellwig wrote:
>> On Thu, Jul 30, 2015 at 05:42:16PM -0400, Doug Ledford wrote:
>>> I have no problem with this code.  That Al finds the user space ABI for
>>> this driver to be bizarre is neither here nor there to me.  Sure, this
>>> file does not exhibit normal file API behavior.  Who cares?
>>
>> Everyone who cares about filesystem semantics does.
>>
>> A NACK from me for a this features, and b) for trying to sneak it in
>> after a negative comment without cc to -fsdevel and Al.
> 
> FWIW, the lack of comments appears to have tripped Doug, judging by his
> "another option would be to drop the write function altogether and just make
> all commands come through writev/write_iter and if you only have one command,
> you only send one element".
> 
> The thing is, ipath and qib (and AFAICS this one as well) have write(2) and
> writev(2) take different and completely unrelated sets of commands.  On
> the same file.  IOW, the effects of
> 	writev(fd, &(struct iovec){buf, len}, 1)
> and
> 	write(fd, buf, len)
> are not even remotely similar.  _That_ is the gratuitous weirdness I'd been
> unhappy with.  And yes, it is gratuitous - it's trivial to have separate files
> for separate command sets.

Yes, admittedly I did think that the command sets were the same.
However, even with them not being the same, I still don't care.  This is
a private driver interface.  Nothing uses it but libpsm or libpsm2, both
of which are the mated user space libraries that implement the user side
of this interface.  It generally isn't intended for anyone else to use.
 If this were a public API, I would care.  It isn't.  All that *really*
matters here is that their kernel driver and user space library speak
the same language.

As for separate files, that presents its own issues, and for little to
no benefit.  Again, this isn't a public API.  It threw you for a loop
because it isn't what you expected.  But that hardly matters, you aren't
writing a user space application to work with the thing.  It is,
however, exactly what the existing library expects.

> If you drop ->write() in there, you certainly lose the weirdness - along with
> one of those command sets.  Sure, having individual iovecs correspond to
> separate datagrams is fine; tons of drivers are like that.  qib and ipath
> are unique, though, in having *two* command sets overloaded on the same file,
> with write() vs. writev() acting as selector (BTW, single-element AIO
> going like writev(), not like write()).

It sounds weird to someone who isn't used to it, but sounds perfectly
fine to someone used to it.  If this were a public API, I would care,
but this is Intel's private communication channel between their kernel
driver and their user space library.  I find it hard to justify telling
them they need to re-engineer both their kernel driver and their library
because a disinterested third party finds their particular means of
organizing their communications "weird".

> PS: I'm back after several weeks of being sick (recipe for fun: start with
> 40C in shade, get completely soaked in serious rain, then have half an hour
> cab ride, with AC set to ~15C) and while I'd managed to get the mailbox
> down to relatively sane size, I might have easily deleted more than I should
> have.  If there had been anything important sent my way and I don't reply to 
> it by Saturday, please resend.
>
Doug Ledford Aug. 1, 2015, 8:18 p.m. UTC | #10
On 07/31/2015 03:31 AM, Christoph Hellwig wrote:
> On Thu, Jul 30, 2015 at 05:42:16PM -0400, Doug Ledford wrote:
>> I have no problem with this code.  That Al finds the user space ABI for
>> this driver to be bizarre is neither here nor there to me.  Sure, this
>> file does not exhibit normal file API behavior.  Who cares?
> 
> Everyone who cares about filesystem semantics does.

If this were a filesystem and this were its file semantics, that would
be all kinds of screwed up.  But this is a private character special device.

> A NACK from me for a this features,

If you have a legitimate technical reason to NACK this feature, make
your case.  I've publicly stated, in response to Al no less, that I
don't see justification for making a team re-engineer a working, private
interface because a disinterested third party finds it "weird".  If you
provide no valid technical justification for your NACK, I'm going to
disregard it.

> and b) for trying to sneak it in
> after a negative comment without cc to -fsdevel and Al.

Not that I was trying to "sneak it in", but that's neither here nor
there, you Cc:ed Al for me, and we already had our discussion (unless Al
wishes to come back and say more, he so far has allowed me to have the
last word).
Christoph Hellwig Aug. 3, 2015, 3:27 p.m. UTC | #11
On Sat, Aug 01, 2015 at 04:18:31PM -0400, Doug Ledford wrote:
> If you have a legitimate technical reason to NACK this feature, make
> your case.  I've publicly stated, in response to Al no less, that I
> don't see justification for making a team re-engineer a working, private
> interface because a disinterested third party finds it "weird".  If you
> provide no valid technical justification for your NACK, I'm going to
> disregard it.

It's breaking the semantics that write("foo", len) is the same as writev
of a single vector using writev.  We had to add a workaround for the
few broken places doing this which we'd like to remove in the long term.

Note that private driver interface really doesn't matter here.
Everything using file operations does go through the VFS so it does have
major implication on the semantics for all the kernel.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/infiniband/hw/hfi1/file_ops.c b/drivers/infiniband/hw/hfi1/file_ops.c
new file mode 100644
index 0000000..4698617
--- /dev/null
+++ b/drivers/infiniband/hw/hfi1/file_ops.c
@@ -0,0 +1,2140 @@ 
+/*
+ *
+ * This file is provided under a dual BSD/GPLv2 license.  When using or
+ * redistributing this file, you may do so under either license.
+ *
+ * GPL LICENSE SUMMARY
+ *
+ * Copyright(c) 2015 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of version 2 of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * General Public License for more details.
+ *
+ * BSD LICENSE
+ *
+ * Copyright(c) 2015 Intel Corporation.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  - Redistributions of source code must retain the above copyright
+ *    notice, this list of conditions and the following disclaimer.
+ *  - Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in
+ *    the documentation and/or other materials provided with the
+ *    distribution.
+ *  - Neither the name of Intel Corporation nor the names of its
+ *    contributors may be used to endorse or promote products derived
+ *    from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ *
+ */
+#include <linux/pci.h>
+#include <linux/poll.h>
+#include <linux/cdev.h>
+#include <linux/swap.h>
+#include <linux/vmalloc.h>
+#include <linux/highmem.h>
+#include <linux/io.h>
+#include <linux/jiffies.h>
+#include <asm/pgtable.h>
+#include <linux/delay.h>
+#include <linux/export.h>
+#include <linux/module.h>
+#include <linux/cred.h>
+#include <linux/uio.h>
+
+#include "hfi.h"
+#include "pio.h"
+#include "device.h"
+#include "common.h"
+#include "trace.h"
+#include "user_sdma.h"
+#include "eprom.h"
+
+#undef pr_fmt
+#define pr_fmt(fmt) DRIVER_NAME ": " fmt
+
+#define SEND_CTXT_HALT_TIMEOUT 1000 /* msecs */
+
+/*
+ * File operation functions
+ */
+static int hfi1_file_open(struct inode *, struct file *);
+static int hfi1_file_close(struct inode *, struct file *);
+static ssize_t hfi1_file_write(struct file *, const char __user *,
+			       size_t, loff_t *);
+static ssize_t hfi1_write_iter(struct kiocb *, struct iov_iter *);
+static unsigned int hfi1_poll(struct file *, struct poll_table_struct *);
+static int hfi1_file_mmap(struct file *, struct vm_area_struct *);
+
+static u64 kvirt_to_phys(void *);
+static int assign_ctxt(struct file *, struct hfi1_user_info *);
+static int init_subctxts(struct hfi1_ctxtdata *, const struct hfi1_user_info *);
+static int user_init(struct file *);
+static int get_ctxt_info(struct file *, void __user *, __u32);
+static int get_base_info(struct file *, void __user *, __u32);
+static int setup_ctxt(struct file *);
+static int setup_subctxt(struct hfi1_ctxtdata *);
+static int get_user_context(struct file *, struct hfi1_user_info *,
+			    int, unsigned);
+static int find_shared_ctxt(struct file *, const struct hfi1_user_info *);
+static int allocate_ctxt(struct file *, struct hfi1_devdata *,
+			 struct hfi1_user_info *);
+static unsigned int poll_urgent(struct file *, struct poll_table_struct *);
+static unsigned int poll_next(struct file *, struct poll_table_struct *);
+static int user_event_ack(struct hfi1_ctxtdata *, int, unsigned long);
+static int set_ctxt_pkey(struct hfi1_ctxtdata *, unsigned, u16);
+static int manage_rcvq(struct hfi1_ctxtdata *, unsigned, int);
+static int vma_fault(struct vm_area_struct *, struct vm_fault *);
+static int exp_tid_setup(struct file *, struct hfi1_tid_info *);
+static int exp_tid_free(struct file *, struct hfi1_tid_info *);
+static void unlock_exp_tids(struct hfi1_ctxtdata *);
+
+static const struct file_operations hfi1_file_ops = {
+	.owner = THIS_MODULE,
+	.write = hfi1_file_write,
+	.write_iter = hfi1_write_iter,
+	.open = hfi1_file_open,
+	.release = hfi1_file_close,
+	.poll = hfi1_poll,
+	.mmap = hfi1_file_mmap,
+	.llseek = noop_llseek,
+};
+
+static struct vm_operations_struct vm_ops = {
+	.fault = vma_fault,
+};
+
+/*
+ * Types of memories mapped into user processes' space
+ */
+enum mmap_types {
+	PIO_BUFS = 1,
+	PIO_BUFS_SOP,
+	PIO_CRED,
+	RCV_HDRQ,
+	RCV_EGRBUF,
+	UREGS,
+	EVENTS,
+	STATUS,
+	RTAIL,
+	SUBCTXT_UREGS,
+	SUBCTXT_RCV_HDRQ,
+	SUBCTXT_EGRBUF,
+	SDMA_COMP
+};
+
+/*
+ * Masks and offsets defining the mmap tokens
+ */
+#define HFI1_MMAP_OFFSET_MASK   0xfffULL
+#define HFI1_MMAP_OFFSET_SHIFT  0
+#define HFI1_MMAP_SUBCTXT_MASK  0xfULL
+#define HFI1_MMAP_SUBCTXT_SHIFT 12
+#define HFI1_MMAP_CTXT_MASK     0xffULL
+#define HFI1_MMAP_CTXT_SHIFT    16
+#define HFI1_MMAP_TYPE_MASK     0xfULL
+#define HFI1_MMAP_TYPE_SHIFT    24
+#define HFI1_MMAP_MAGIC_MASK    0xffffffffULL
+#define HFI1_MMAP_MAGIC_SHIFT   32
+
+#define HFI1_MMAP_MAGIC         0xdabbad00
+
+#define HFI1_MMAP_TOKEN_SET(field, val)	\
+	(((val) & HFI1_MMAP_##field##_MASK) << HFI1_MMAP_##field##_SHIFT)
+#define HFI1_MMAP_TOKEN_GET(field, token) \
+	(((token) >> HFI1_MMAP_##field##_SHIFT) & HFI1_MMAP_##field##_MASK)
+#define HFI1_MMAP_TOKEN(type, ctxt, subctxt, addr)   \
+	(HFI1_MMAP_TOKEN_SET(MAGIC, HFI1_MMAP_MAGIC) | \
+	HFI1_MMAP_TOKEN_SET(TYPE, type) | \
+	HFI1_MMAP_TOKEN_SET(CTXT, ctxt) | \
+	HFI1_MMAP_TOKEN_SET(SUBCTXT, subctxt) | \
+	HFI1_MMAP_TOKEN_SET(OFFSET, ((unsigned long)addr & ~PAGE_MASK)))
+
+#define EXP_TID_SET(field, value)			\
+	(((value) & EXP_TID_TID##field##_MASK) <<	\
+	 EXP_TID_TID##field##_SHIFT)
+#define EXP_TID_CLEAR(tid, field) {					\
+		(tid) &= ~(EXP_TID_TID##field##_MASK <<			\
+			   EXP_TID_TID##field##_SHIFT);			\
+			}
+#define EXP_TID_RESET(tid, field, value) do {				\
+		EXP_TID_CLEAR(tid, field);				\
+		(tid) |= EXP_TID_SET(field, value);			\
+	} while (0)
+
+#define dbg(fmt, ...)				\
+	pr_info(fmt, ##__VA_ARGS__)
+
+
+static inline int is_valid_mmap(u64 token)
+{
+	return (HFI1_MMAP_TOKEN_GET(MAGIC, token) == HFI1_MMAP_MAGIC);
+}
+
+static int hfi1_file_open(struct inode *inode, struct file *fp)
+{
+	/* The real work is performed later in assign_ctxt() */
+	fp->private_data = kzalloc(sizeof(struct hfi1_filedata), GFP_KERNEL);
+	if (fp->private_data) /* no cpu affinity by default */
+		((struct hfi1_filedata *)fp->private_data)->rec_cpu_num = -1;
+	return fp->private_data ? 0 : -ENOMEM;
+}
+
+static ssize_t hfi1_file_write(struct file *fp, const char __user *data,
+			       size_t count, loff_t *offset)
+{
+	const struct hfi1_cmd __user *ucmd;
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_cmd cmd;
+	struct hfi1_user_info uinfo;
+	struct hfi1_tid_info tinfo;
+	ssize_t consumed = 0, copy = 0, ret = 0;
+	void *dest = NULL;
+	__u64 user_val = 0;
+	int uctxt_required = 1;
+	int must_be_root = 0;
+
+	if (count < sizeof(cmd)) {
+		ret = -EINVAL;
+		goto bail;
+	}
+
+	ucmd = (const struct hfi1_cmd __user *)data;
+	if (copy_from_user(&cmd, ucmd, sizeof(cmd))) {
+		ret = -EFAULT;
+		goto bail;
+	}
+
+	consumed = sizeof(cmd);
+
+	switch (cmd.type) {
+	case HFI1_CMD_ASSIGN_CTXT:
+		uctxt_required = 0;	/* assigned user context not required */
+		copy = sizeof(uinfo);
+		dest = &uinfo;
+		break;
+	case HFI1_CMD_SDMA_STATUS_UPD:
+	case HFI1_CMD_CREDIT_UPD:
+		copy = 0;
+		break;
+	case HFI1_CMD_TID_UPDATE:
+	case HFI1_CMD_TID_FREE:
+		copy = sizeof(tinfo);
+		dest = &tinfo;
+		break;
+	case HFI1_CMD_USER_INFO:
+	case HFI1_CMD_RECV_CTRL:
+	case HFI1_CMD_POLL_TYPE:
+	case HFI1_CMD_ACK_EVENT:
+	case HFI1_CMD_CTXT_INFO:
+	case HFI1_CMD_SET_PKEY:
+	case HFI1_CMD_CTXT_RESET:
+		copy = 0;
+		user_val = cmd.addr;
+		break;
+	case HFI1_CMD_EP_INFO:
+	case HFI1_CMD_EP_ERASE_CHIP:
+	case HFI1_CMD_EP_ERASE_P0:
+	case HFI1_CMD_EP_ERASE_P1:
+	case HFI1_CMD_EP_READ_P0:
+	case HFI1_CMD_EP_READ_P1:
+	case HFI1_CMD_EP_WRITE_P0:
+	case HFI1_CMD_EP_WRITE_P1:
+		uctxt_required = 0;	/* assigned user context not required */
+		must_be_root = 1;	/* validate user */
+		copy = 0;
+		break;
+	default:
+		ret = -EINVAL;
+		goto bail;
+	}
+
+	/* If the command comes with user data, copy it. */
+	if (copy) {
+		if (copy_from_user(dest, (void __user *)cmd.addr, copy)) {
+			ret = -EFAULT;
+			goto bail;
+		}
+		consumed += copy;
+	}
+
+	/*
+	 * Make sure there is a uctxt when needed.
+	 */
+	if (uctxt_required && !uctxt) {
+		ret = -EINVAL;
+		goto bail;
+	}
+
+	/* only root can do these operations */
+	if (must_be_root && !capable(CAP_SYS_ADMIN)) {
+		ret = -EPERM;
+		goto bail;
+	}
+
+	switch (cmd.type) {
+	case HFI1_CMD_ASSIGN_CTXT:
+		ret = assign_ctxt(fp, &uinfo);
+		if (ret < 0)
+			goto bail;
+		ret = setup_ctxt(fp);
+		if (ret)
+			goto bail;
+		ret = user_init(fp);
+		break;
+	case HFI1_CMD_CTXT_INFO:
+		ret = get_ctxt_info(fp, (void __user *)(unsigned long)
+				    user_val, cmd.len);
+		break;
+	case HFI1_CMD_USER_INFO:
+		ret = get_base_info(fp, (void __user *)(unsigned long)
+				    user_val, cmd.len);
+		break;
+	case HFI1_CMD_SDMA_STATUS_UPD:
+		break;
+	case HFI1_CMD_CREDIT_UPD:
+		if (uctxt && uctxt->sc)
+			sc_return_credits(uctxt->sc);
+		break;
+	case HFI1_CMD_TID_UPDATE:
+		ret = exp_tid_setup(fp, &tinfo);
+		if (!ret) {
+			unsigned long addr;
+			/*
+			 * Copy the number of tidlist entries we used
+			 * and the length of the buffer we registered.
+			 * These fields are adjacent in the structure so
+			 * we can copy them at the same time.
+			 */
+			addr = (unsigned long)cmd.addr +
+				offsetof(struct hfi1_tid_info, tidcnt);
+			if (copy_to_user((void __user *)addr, &tinfo.tidcnt,
+					 sizeof(tinfo.tidcnt) +
+					 sizeof(tinfo.length)))
+				ret = -EFAULT;
+		}
+		break;
+	case HFI1_CMD_TID_FREE:
+		ret = exp_tid_free(fp, &tinfo);
+		break;
+	case HFI1_CMD_RECV_CTRL:
+		ret = manage_rcvq(uctxt, subctxt_fp(fp), (int)user_val);
+		break;
+	case HFI1_CMD_POLL_TYPE:
+		uctxt->poll_type = (typeof(uctxt->poll_type))user_val;
+		break;
+	case HFI1_CMD_ACK_EVENT:
+		ret = user_event_ack(uctxt, subctxt_fp(fp), user_val);
+		break;
+	case HFI1_CMD_SET_PKEY:
+		if (HFI1_CAP_IS_USET(PKEY_CHECK))
+			ret = set_ctxt_pkey(uctxt, subctxt_fp(fp), user_val);
+		else
+			ret = -EPERM;
+		break;
+	case HFI1_CMD_CTXT_RESET: {
+		struct send_context *sc;
+		struct hfi1_devdata *dd;
+
+		if (!uctxt || !uctxt->dd || !uctxt->sc) {
+			ret = -EINVAL;
+			break;
+		}
+		/*
+		 * There is no protection here. User level has to
+		 * guarantee that no one will be writing to the send
+		 * context while it is being re-initialized.
+		 * If user level breaks that guarantee, it will break
+		 * it's own context and no one else's.
+		 */
+		dd = uctxt->dd;
+		sc = uctxt->sc;
+		/*
+		 * Wait until the interrupt handler has marked the
+		 * context as halted or frozen. Report error if we time
+		 * out.
+		 */
+		wait_event_interruptible_timeout(
+			sc->halt_wait, (sc->flags & SCF_HALTED),
+			msecs_to_jiffies(SEND_CTXT_HALT_TIMEOUT));
+		if (!(sc->flags & SCF_HALTED)) {
+			ret = -ENOLCK;
+			break;
+		}
+		/*
+		 * If the send context was halted due to a Freeze,
+		 * wait until the device has been "unfrozen" before
+		 * resetting the context.
+		 */
+		if (sc->flags & SCF_FROZEN) {
+			wait_event_interruptible_timeout(
+				dd->event_queue,
+				!(ACCESS_ONCE(dd->flags) & HFI1_FROZEN),
+				msecs_to_jiffies(SEND_CTXT_HALT_TIMEOUT));
+			if (dd->flags & HFI1_FROZEN) {
+				ret = -ENOLCK;
+				break;
+			}
+			if (dd->flags & HFI1_FORCED_FREEZE) {
+				/* Don't allow context reset if we are into
+				 * forced freeze */
+				ret = -ENODEV;
+				break;
+			}
+			sc_disable(sc);
+			ret = sc_enable(sc);
+			hfi1_rcvctrl(dd, HFI1_RCVCTRL_CTXT_ENB,
+				     uctxt->ctxt);
+		} else
+			ret = sc_restart(sc);
+		if (!ret)
+			sc_return_credits(sc);
+		break;
+	}
+	case HFI1_CMD_EP_INFO:
+	case HFI1_CMD_EP_ERASE_CHIP:
+	case HFI1_CMD_EP_ERASE_P0:
+	case HFI1_CMD_EP_ERASE_P1:
+	case HFI1_CMD_EP_READ_P0:
+	case HFI1_CMD_EP_READ_P1:
+	case HFI1_CMD_EP_WRITE_P0:
+	case HFI1_CMD_EP_WRITE_P1:
+		ret = handle_eprom_command(&cmd);
+		break;
+	}
+
+	if (ret >= 0)
+		ret = consumed;
+bail:
+	return ret;
+}
+
+static ssize_t hfi1_write_iter(struct kiocb *kiocb, struct iov_iter *from)
+{
+	struct hfi1_user_sdma_pkt_q *pq;
+	struct hfi1_user_sdma_comp_q *cq;
+	int ret = 0, done = 0, reqs = 0;
+	unsigned long dim = from->nr_segs;
+
+	if (!user_sdma_comp_fp(kiocb->ki_filp) ||
+	    !user_sdma_pkt_fp(kiocb->ki_filp)) {
+		ret = -EIO;
+		goto done;
+	}
+
+	if (!iter_is_iovec(from) || !dim) {
+		ret = -EINVAL;
+		goto done;
+	}
+
+	hfi1_cdbg(SDMA, "SDMA request from %u:%u (%lu)",
+		  ctxt_fp(kiocb->ki_filp)->ctxt, subctxt_fp(kiocb->ki_filp),
+		  dim);
+	pq = user_sdma_pkt_fp(kiocb->ki_filp);
+	cq = user_sdma_comp_fp(kiocb->ki_filp);
+
+	if (atomic_read(&pq->n_reqs) == pq->n_max_reqs) {
+		ret = -ENOSPC;
+		goto done;
+	}
+
+	while (dim) {
+		unsigned long count = 0;
+
+		ret = hfi1_user_sdma_process_request(
+			kiocb->ki_filp,	(struct iovec *)(from->iov + done),
+			dim, &count);
+		if (ret)
+			goto done;
+		dim -= count;
+		done += count;
+		reqs++;
+	}
+done:
+	return ret ? ret : reqs;
+}
+
+static int hfi1_file_mmap(struct file *fp, struct vm_area_struct *vma)
+{
+	struct hfi1_ctxtdata *uctxt;
+	struct hfi1_devdata *dd;
+	unsigned long flags, pfn;
+	u64 token = vma->vm_pgoff << PAGE_SHIFT,
+		memaddr = 0;
+	u8 subctxt, mapio = 0, vmf = 0, type;
+	ssize_t memlen = 0;
+	int ret = 0;
+	u16 ctxt;
+
+	uctxt = ctxt_fp(fp);
+	if (!is_valid_mmap(token) || !uctxt ||
+	    !(vma->vm_flags & VM_SHARED)) {
+		ret = -EINVAL;
+		goto done;
+	}
+	dd = uctxt->dd;
+	ctxt = HFI1_MMAP_TOKEN_GET(CTXT, token);
+	subctxt = HFI1_MMAP_TOKEN_GET(SUBCTXT, token);
+	type = HFI1_MMAP_TOKEN_GET(TYPE, token);
+	if (ctxt != uctxt->ctxt || subctxt != subctxt_fp(fp)) {
+		ret = -EINVAL;
+		goto done;
+	}
+
+	flags = vma->vm_flags;
+
+	switch (type) {
+	case PIO_BUFS:
+	case PIO_BUFS_SOP:
+		memaddr = ((dd->physaddr + TXE_PIO_SEND) +
+				/* chip pio base */
+			   (uctxt->sc->hw_context * (1 << 16))) +
+				/* 64K PIO space / ctxt */
+			(type == PIO_BUFS_SOP ?
+				(TXE_PIO_SIZE / 2) : 0); /* sop? */
+		/*
+		 * Map only the amount allocated to the context, not the
+		 * entire available context's PIO space.
+		 */
+		memlen = ALIGN(uctxt->sc->credits * PIO_BLOCK_SIZE,
+			       PAGE_SIZE);
+		flags &= ~VM_MAYREAD;
+		flags |= VM_DONTCOPY | VM_DONTEXPAND;
+		vma->vm_page_prot = pgprot_writecombine(vma->vm_page_prot);
+		mapio = 1;
+		break;
+	case PIO_CRED:
+		if (flags & VM_WRITE) {
+			ret = -EPERM;
+			goto done;
+		}
+		/*
+		 * The credit return location for this context could be on the
+		 * second or third page allocated for credit returns (if number
+		 * of enabled contexts > 64 and 128 respectively).
+		 */
+		memaddr = dd->cr_base[uctxt->numa_id].pa +
+			(((u64)uctxt->sc->hw_free -
+			  (u64)dd->cr_base[uctxt->numa_id].va) & PAGE_MASK);
+		memlen = PAGE_SIZE;
+		flags &= ~VM_MAYWRITE;
+		flags |= VM_DONTCOPY | VM_DONTEXPAND;
+		/*
+		 * The driver has already allocated memory for credit
+		 * returns and programmed it into the chip. Has that
+		 * memory been flagged as non-cached?
+		 */
+		/* vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); */
+		mapio = 1;
+		break;
+	case RCV_HDRQ:
+		memaddr = uctxt->rcvhdrq_phys;
+		memlen = uctxt->rcvhdrq_size;
+		break;
+	case RCV_EGRBUF: {
+		unsigned long addr;
+		int i;
+		/*
+		 * The RcvEgr buffer need to be handled differently
+		 * as multiple non-contiguous pages need to be mapped
+		 * into the user process.
+		 */
+		memlen = uctxt->egrbufs.size;
+		if ((vma->vm_end - vma->vm_start) != memlen) {
+			dd_dev_err(dd, "Eager buffer map size invalid (%lu != %lu)\n",
+				   (vma->vm_end - vma->vm_start), memlen);
+			ret = -EINVAL;
+			goto done;
+		}
+		if (vma->vm_flags & VM_WRITE) {
+			ret = -EPERM;
+			goto done;
+		}
+		vma->vm_flags &= ~VM_MAYWRITE;
+		addr = vma->vm_start;
+		for (i = 0 ; i < uctxt->egrbufs.numbufs; i++) {
+			ret = remap_pfn_range(
+				vma, addr,
+				uctxt->egrbufs.buffers[i].phys >> PAGE_SHIFT,
+				uctxt->egrbufs.buffers[i].len,
+				vma->vm_page_prot);
+			if (ret < 0)
+				goto done;
+			addr += uctxt->egrbufs.buffers[i].len;
+		}
+		ret = 0;
+		goto done;
+	}
+	case UREGS:
+		/*
+		 * Map only the page that contains this context's user
+		 * registers.
+		 */
+		memaddr = (unsigned long)
+			(dd->physaddr + RXE_PER_CONTEXT_USER)
+			+ (uctxt->ctxt * RXE_PER_CONTEXT_SIZE);
+		/*
+		 * TidFlow table is on the same page as the rest of the
+		 * user registers.
+		 */
+		memlen = PAGE_SIZE;
+		flags |= VM_DONTCOPY | VM_DONTEXPAND;
+		vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
+		mapio = 1;
+		break;
+	case EVENTS:
+		/*
+		 * Use the page where this context's flags are. User level
+		 * knows where it's own bitmap is within the page.
+		 */
+		memaddr = ((unsigned long)dd->events +
+			   ((uctxt->ctxt - dd->first_user_ctxt) *
+			    HFI1_MAX_SHARED_CTXTS)) & PAGE_MASK;
+		memlen = PAGE_SIZE;
+		/*
+		 * v3.7 removes VM_RESERVED but the effect is kept by
+		 * using VM_IO.
+		 */
+		flags |= VM_IO | VM_DONTEXPAND;
+		vmf = 1;
+		break;
+	case STATUS:
+		memaddr = kvirt_to_phys((void *)dd->status);
+		memlen = PAGE_SIZE;
+		flags |= VM_IO | VM_DONTEXPAND;
+		break;
+	case RTAIL:
+		if (!HFI1_CAP_IS_USET(DMA_RTAIL)) {
+			/*
+			 * If the memory allocation failed, the context alloc
+			 * also would have failed, so we would never get here
+			 */
+			ret = -EINVAL;
+			goto done;
+		}
+		if (flags & VM_WRITE) {
+			ret = -EPERM;
+			goto done;
+		}
+		memaddr = uctxt->rcvhdrqtailaddr_phys;
+		memlen = PAGE_SIZE;
+		flags &= ~VM_MAYWRITE;
+		break;
+	case SUBCTXT_UREGS:
+		memaddr = (u64)uctxt->subctxt_uregbase;
+		memlen = PAGE_SIZE;
+		flags |= VM_IO | VM_DONTEXPAND;
+		vmf = 1;
+		break;
+	case SUBCTXT_RCV_HDRQ:
+		memaddr = (u64)uctxt->subctxt_rcvhdr_base;
+		memlen = uctxt->rcvhdrq_size * uctxt->subctxt_cnt;
+		flags |= VM_IO | VM_DONTEXPAND;
+		vmf = 1;
+		break;
+	case SUBCTXT_EGRBUF:
+		memaddr = (u64)uctxt->subctxt_rcvegrbuf;
+		memlen = uctxt->egrbufs.size * uctxt->subctxt_cnt;
+		flags |= VM_IO | VM_DONTEXPAND;
+		flags &= ~VM_MAYWRITE;
+		vmf = 1;
+		break;
+	case SDMA_COMP: {
+		struct hfi1_user_sdma_comp_q *cq;
+
+		if (!user_sdma_comp_fp(fp)) {
+			ret = -EFAULT;
+			goto done;
+		}
+		cq = user_sdma_comp_fp(fp);
+		memaddr = (u64)cq->comps;
+		memlen = ALIGN(sizeof(*cq->comps) * cq->nentries, PAGE_SIZE);
+		flags |= VM_IO | VM_DONTEXPAND;
+		vmf = 1;
+		break;
+	}
+	default:
+		ret = -EINVAL;
+		break;
+	}
+
+	if ((vma->vm_end - vma->vm_start) != memlen) {
+		hfi1_cdbg(PROC, "%u:%u Memory size mismatch %lu:%lu",
+			  uctxt->ctxt, subctxt_fp(fp),
+			  (vma->vm_end - vma->vm_start), memlen);
+		ret = -EINVAL;
+		goto done;
+	}
+
+	vma->vm_flags = flags;
+	dd_dev_info(dd,
+		    "%s: %u:%u type:%u io/vf:%d/%d, addr:0x%llx, len:%lu(%lu), flags:0x%lx\n",
+		    __func__, ctxt, subctxt, type, mapio, vmf, memaddr, memlen,
+		    vma->vm_end - vma->vm_start, vma->vm_flags);
+	pfn = (unsigned long)(memaddr >> PAGE_SHIFT);
+	if (vmf) {
+		vma->vm_pgoff = pfn;
+		vma->vm_ops = &vm_ops;
+		ret = 0;
+	} else if (mapio) {
+		ret = io_remap_pfn_range(vma, vma->vm_start, pfn, memlen,
+					 vma->vm_page_prot);
+	} else {
+		ret = remap_pfn_range(vma, vma->vm_start, pfn, memlen,
+				      vma->vm_page_prot);
+	}
+done:
+	return ret;
+}
+
+/*
+ * Local (non-chip) user memory is not mapped right away but as it is
+ * accessed by the user-level code.
+ */
+static int vma_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
+{
+	struct page *page;
+
+	page = vmalloc_to_page((void *)(vmf->pgoff << PAGE_SHIFT));
+	if (!page)
+		return VM_FAULT_SIGBUS;
+
+	get_page(page);
+	vmf->page = page;
+
+	return 0;
+}
+
+static unsigned int hfi1_poll(struct file *fp, struct poll_table_struct *pt)
+{
+	struct hfi1_ctxtdata *uctxt;
+	unsigned pollflag;
+
+	uctxt = ctxt_fp(fp);
+	if (!uctxt)
+		pollflag = POLLERR;
+	else if (uctxt->poll_type == HFI1_POLL_TYPE_URGENT)
+		pollflag = poll_urgent(fp, pt);
+	else  if (uctxt->poll_type == HFI1_POLL_TYPE_ANYRCV)
+		pollflag = poll_next(fp, pt);
+	else /* invalid */
+		pollflag = POLLERR;
+
+	return pollflag;
+}
+
+static int hfi1_file_close(struct inode *inode, struct file *fp)
+{
+	struct hfi1_filedata *fdata = fp->private_data;
+	struct hfi1_ctxtdata *uctxt = fdata->uctxt;
+	struct hfi1_devdata *dd;
+	unsigned long flags, *ev;
+
+	fp->private_data = NULL;
+
+	if (!uctxt)
+		goto done;
+
+	hfi1_cdbg(PROC, "freeing ctxt %u:%u", uctxt->ctxt, fdata->subctxt);
+	dd = uctxt->dd;
+	mutex_lock(&hfi1_mutex);
+
+	flush_wc();
+	/* drain user sdma queue */
+	if (fdata->pq)
+		hfi1_user_sdma_free_queues(fdata);
+
+	/*
+	 * Clear any left over, unhandled events so the next process that
+	 * gets this context doesn't get confused.
+	 */
+	ev = dd->events + ((uctxt->ctxt - dd->first_user_ctxt) *
+			   HFI1_MAX_SHARED_CTXTS) + fdata->subctxt;
+	*ev = 0;
+
+	if (--uctxt->cnt) {
+		uctxt->active_slaves &= ~(1 << fdata->subctxt);
+		uctxt->subpid[fdata->subctxt] = 0;
+		mutex_unlock(&hfi1_mutex);
+		goto done;
+	}
+
+	spin_lock_irqsave(&dd->uctxt_lock, flags);
+	/*
+	 * Disable receive context and interrupt available, reset all
+	 * RcvCtxtCtrl bits to default values.
+	 */
+	hfi1_rcvctrl(dd, HFI1_RCVCTRL_CTXT_DIS |
+		     HFI1_RCVCTRL_TIDFLOW_DIS |
+		     HFI1_RCVCTRL_INTRAVAIL_DIS |
+		     HFI1_RCVCTRL_ONE_PKT_EGR_DIS |
+		     HFI1_RCVCTRL_NO_RHQ_DROP_DIS |
+		     HFI1_RCVCTRL_NO_EGR_DROP_DIS, uctxt->ctxt);
+	/* Clear the context's J_KEY */
+	hfi1_clear_ctxt_jkey(dd, uctxt->ctxt);
+	/*
+	 * Reset context integrity checks to default.
+	 * (writes to CSRs probably belong in chip.c)
+	 */
+	write_kctxt_csr(dd, uctxt->sc->hw_context, SEND_CTXT_CHECK_ENABLE,
+			hfi1_pkt_default_send_ctxt_mask(dd, uctxt->sc->type));
+	sc_disable(uctxt->sc);
+	uctxt->pid = 0;
+	spin_unlock_irqrestore(&dd->uctxt_lock, flags);
+
+	dd->rcd[uctxt->ctxt] = NULL;
+	uctxt->rcvwait_to = 0;
+	uctxt->piowait_to = 0;
+	uctxt->rcvnowait = 0;
+	uctxt->pionowait = 0;
+	uctxt->event_flags = 0;
+
+	hfi1_clear_tids(uctxt);
+	hfi1_clear_ctxt_pkey(dd, uctxt->ctxt);
+
+	if (uctxt->tid_pg_list)
+		unlock_exp_tids(uctxt);
+
+	hfi1_stats.sps_ctxts--;
+	dd->freectxts++;
+	mutex_unlock(&hfi1_mutex);
+	hfi1_free_ctxtdata(dd, uctxt);
+done:
+	kfree(fdata);
+	return 0;
+}
+
+/*
+ * Convert kernel *virtual* addresses to physical addresses.
+ * This is used to vmalloc'ed addresses.
+ */
+static u64 kvirt_to_phys(void *addr)
+{
+	struct page *page;
+	u64 paddr = 0;
+
+	page = vmalloc_to_page(addr);
+	if (page)
+		paddr = page_to_pfn(page) << PAGE_SHIFT;
+
+	return paddr;
+}
+
+static int assign_ctxt(struct file *fp, struct hfi1_user_info *uinfo)
+{
+	int i_minor, ret = 0;
+	unsigned swmajor, swminor, alg = HFI1_ALG_ACROSS;
+
+	swmajor = uinfo->userversion >> 16;
+	if (swmajor != HFI1_USER_SWMAJOR) {
+		ret = -ENODEV;
+		goto done;
+	}
+
+	swminor = uinfo->userversion & 0xffff;
+
+	if (uinfo->hfi1_alg < HFI1_ALG_COUNT)
+		alg = uinfo->hfi1_alg;
+
+	mutex_lock(&hfi1_mutex);
+	/* First, lets check if we need to setup a shared context? */
+	if (uinfo->subctxt_cnt)
+		ret = find_shared_ctxt(fp, uinfo);
+
+	/*
+	 * We execute the following block if we couldn't find a
+	 * shared context or if context sharing is not required.
+	 */
+	if (!ret) {
+		i_minor = iminor(file_inode(fp)) - HFI1_USER_MINOR_BASE;
+		ret = get_user_context(fp, uinfo, i_minor - 1, alg);
+	}
+	mutex_unlock(&hfi1_mutex);
+done:
+	return ret;
+}
+
+static int get_user_context(struct file *fp, struct hfi1_user_info *uinfo,
+			    int devno, unsigned alg)
+{
+	struct hfi1_devdata *dd = NULL;
+	int ret = 0, devmax, npresent, nup, dev;
+
+	devmax = hfi1_count_units(&npresent, &nup);
+	if (!npresent) {
+		ret = -ENXIO;
+		goto done;
+	}
+	if (!nup) {
+		ret = -ENETDOWN;
+		goto done;
+	}
+	if (devno >= 0) {
+		dd = hfi1_lookup(devno);
+		if (!dd)
+			ret = -ENODEV;
+		else if (!dd->freectxts)
+			ret = -EBUSY;
+	} else {
+		struct hfi1_devdata *pdd;
+
+		if (alg == HFI1_ALG_ACROSS) {
+			unsigned free = 0U;
+
+			for (dev = 0; dev < devmax; dev++) {
+				pdd = hfi1_lookup(dev);
+				if (pdd && pdd->freectxts &&
+				    pdd->freectxts > free) {
+					dd = pdd;
+					free = pdd->freectxts;
+				}
+			}
+		} else {
+			for (dev = 0; dev < devmax; dev++) {
+				pdd = hfi1_lookup(dev);
+				if (pdd && pdd->freectxts) {
+					dd = pdd;
+					break;
+				}
+			}
+		}
+		if (!dd)
+			ret = -EBUSY;
+	}
+done:
+	return ret ? ret : allocate_ctxt(fp, dd, uinfo);
+}
+
+static int find_shared_ctxt(struct file *fp,
+			    const struct hfi1_user_info *uinfo)
+{
+	int devmax, ndev, i;
+	int ret = 0;
+
+	devmax = hfi1_count_units(NULL, NULL);
+
+	for (ndev = 0; ndev < devmax; ndev++) {
+		struct hfi1_devdata *dd = hfi1_lookup(ndev);
+
+		/* device portion of usable() */
+		if (!(dd && (dd->flags & HFI1_PRESENT) && dd->kregbase))
+			continue;
+		for (i = dd->first_user_ctxt; i < dd->num_rcv_contexts; i++) {
+			struct hfi1_ctxtdata *uctxt = dd->rcd[i];
+
+			/* Skip ctxts which are not yet open */
+			if (!uctxt || !uctxt->cnt)
+				continue;
+			/* Skip ctxt if it doesn't match the requested one */
+			if (memcmp(uctxt->uuid, uinfo->uuid,
+				   sizeof(uctxt->uuid)) ||
+			    uctxt->subctxt_id != uinfo->subctxt_id ||
+			    uctxt->subctxt_cnt != uinfo->subctxt_cnt)
+				continue;
+
+			/* Verify the sharing process matches the master */
+			if (uctxt->userversion != uinfo->userversion ||
+			    uctxt->cnt >= uctxt->subctxt_cnt) {
+				ret = -EINVAL;
+				goto done;
+			}
+			ctxt_fp(fp) = uctxt;
+			subctxt_fp(fp) = uctxt->cnt++;
+			uctxt->subpid[subctxt_fp(fp)] = current->pid;
+			uctxt->active_slaves |= 1 << subctxt_fp(fp);
+			ret = 1;
+			goto done;
+		}
+	}
+
+done:
+	return ret;
+}
+
+static int allocate_ctxt(struct file *fp, struct hfi1_devdata *dd,
+			 struct hfi1_user_info *uinfo)
+{
+	struct hfi1_ctxtdata *uctxt;
+	unsigned ctxt;
+	int ret;
+
+	if (dd->flags & HFI1_FROZEN) {
+		/*
+		 * Pick an error that is unique from all other errors
+		 * that are returned so the user process knows that
+		 * it tried to allocate while the SPC was frozen.  It
+		 * it should be able to retry with success in a short
+		 * while.
+		 */
+		return -EIO;
+	}
+
+	for (ctxt = dd->first_user_ctxt; ctxt < dd->num_rcv_contexts; ctxt++)
+		if (!dd->rcd[ctxt])
+			break;
+
+	if (ctxt == dd->num_rcv_contexts)
+		return -EBUSY;
+
+	uctxt = hfi1_create_ctxtdata(dd->pport, ctxt);
+	if (!uctxt) {
+		dd_dev_err(dd,
+			   "Unable to allocate ctxtdata memory, failing open\n");
+		return -ENOMEM;
+	}
+	/*
+	 * Allocate and enable a PIO send context.
+	 */
+	uctxt->sc = sc_alloc(dd, SC_USER, uctxt->rcvhdrqentsize,
+			     uctxt->numa_id);
+	if (!uctxt->sc)
+		return -ENOMEM;
+
+	dbg("allocated send context %u(%u)\n", uctxt->sc->sw_index,
+		uctxt->sc->hw_context);
+	ret = sc_enable(uctxt->sc);
+	if (ret)
+		return ret;
+	/*
+	 * Setup shared context resources if the user-level has requested
+	 * shared contexts and this is the 'master' process.
+	 * This has to be done here so the rest of the sub-contexts find the
+	 * proper master.
+	 */
+	if (uinfo->subctxt_cnt && !subctxt_fp(fp)) {
+		ret = init_subctxts(uctxt, uinfo);
+		/*
+		 * On error, we don't need to disable and de-allocate the
+		 * send context because it will be done during file close
+		 */
+		if (ret)
+			return ret;
+	}
+	uctxt->userversion = uinfo->userversion;
+	uctxt->pid = current->pid;
+	uctxt->flags = HFI1_CAP_UGET(MASK);
+	init_waitqueue_head(&uctxt->wait);
+	strlcpy(uctxt->comm, current->comm, sizeof(uctxt->comm));
+	memcpy(uctxt->uuid, uinfo->uuid, sizeof(uctxt->uuid));
+	uctxt->jkey = generate_jkey(current_uid());
+	INIT_LIST_HEAD(&uctxt->sdma_queues);
+	spin_lock_init(&uctxt->sdma_qlock);
+	hfi1_stats.sps_ctxts++;
+	dd->freectxts--;
+	ctxt_fp(fp) = uctxt;
+
+	return 0;
+}
+
+static int init_subctxts(struct hfi1_ctxtdata *uctxt,
+			 const struct hfi1_user_info *uinfo)
+{
+	int ret = 0;
+	unsigned num_subctxts;
+
+	num_subctxts = uinfo->subctxt_cnt;
+	if (num_subctxts > HFI1_MAX_SHARED_CTXTS) {
+		ret = -EINVAL;
+		goto bail;
+	}
+
+	uctxt->subctxt_cnt = uinfo->subctxt_cnt;
+	uctxt->subctxt_id = uinfo->subctxt_id;
+	uctxt->active_slaves = 1;
+	uctxt->redirect_seq_cnt = 1;
+	set_bit(HFI1_CTXT_MASTER_UNINIT, &uctxt->event_flags);
+bail:
+	return ret;
+}
+
+static int setup_subctxt(struct hfi1_ctxtdata *uctxt)
+{
+	int ret = 0;
+	unsigned num_subctxts = uctxt->subctxt_cnt;
+
+	uctxt->subctxt_uregbase = vmalloc_user(PAGE_SIZE);
+	if (!uctxt->subctxt_uregbase) {
+		ret = -ENOMEM;
+		goto bail;
+	}
+	/* We can take the size of the RcvHdr Queue from the master */
+	uctxt->subctxt_rcvhdr_base = vmalloc_user(uctxt->rcvhdrq_size *
+						  num_subctxts);
+	if (!uctxt->subctxt_rcvhdr_base) {
+		ret = -ENOMEM;
+		goto bail_ureg;
+	}
+
+	uctxt->subctxt_rcvegrbuf = vmalloc_user(uctxt->egrbufs.size *
+						num_subctxts);
+	if (!uctxt->subctxt_rcvegrbuf) {
+		ret = -ENOMEM;
+		goto bail_rhdr;
+	}
+	goto bail;
+bail_rhdr:
+	vfree(uctxt->subctxt_rcvhdr_base);
+bail_ureg:
+	vfree(uctxt->subctxt_uregbase);
+	uctxt->subctxt_uregbase = NULL;
+bail:
+	return ret;
+}
+
+static int user_init(struct file *fp)
+{
+	int ret;
+	unsigned int rcvctrl_ops = 0;
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+
+	/* make sure that the context has already been setup */
+	if (!test_bit(HFI1_CTXT_SETUP_DONE, &uctxt->event_flags)) {
+		ret = -EFAULT;
+		goto done;
+	}
+
+	/*
+	 * Subctxts don't need to initialize anything since master
+	 * has done it.
+	 */
+	if (subctxt_fp(fp)) {
+		ret = wait_event_interruptible(uctxt->wait,
+			!test_bit(HFI1_CTXT_MASTER_UNINIT,
+			&uctxt->event_flags));
+		goto done;
+	}
+
+	/* initialize poll variables... */
+	uctxt->urgent = 0;
+	uctxt->urgent_poll = 0;
+
+	/*
+	 * Now enable the ctxt for receive.
+	 * For chips that are set to DMA the tail register to memory
+	 * when they change (and when the update bit transitions from
+	 * 0 to 1.  So for those chips, we turn it off and then back on.
+	 * This will (very briefly) affect any other open ctxts, but the
+	 * duration is very short, and therefore isn't an issue.  We
+	 * explicitly set the in-memory tail copy to 0 beforehand, so we
+	 * don't have to wait to be sure the DMA update has happened
+	 * (chip resets head/tail to 0 on transition to enable).
+	 */
+	if (uctxt->rcvhdrtail_kvaddr)
+		clear_rcvhdrtail(uctxt);
+
+	/* Setup J_KEY before enabling the context */
+	hfi1_set_ctxt_jkey(uctxt->dd, uctxt->ctxt, uctxt->jkey);
+
+	rcvctrl_ops = HFI1_RCVCTRL_CTXT_ENB;
+	if (HFI1_CAP_KGET_MASK(uctxt->flags, HDRSUPP))
+		rcvctrl_ops |= HFI1_RCVCTRL_TIDFLOW_ENB;
+	/*
+	 * Ignore the bit in the flags for now until proper
+	 * support for multiple packet per rcv array entry is
+	 * added.
+	 */
+	if (!HFI1_CAP_KGET_MASK(uctxt->flags, MULTI_PKT_EGR))
+		rcvctrl_ops |= HFI1_RCVCTRL_ONE_PKT_EGR_ENB;
+	if (HFI1_CAP_KGET_MASK(uctxt->flags, NODROP_EGR_FULL))
+		rcvctrl_ops |= HFI1_RCVCTRL_NO_EGR_DROP_ENB;
+	if (HFI1_CAP_KGET_MASK(uctxt->flags, NODROP_RHQ_FULL))
+		rcvctrl_ops |= HFI1_RCVCTRL_NO_RHQ_DROP_ENB;
+	if (HFI1_CAP_KGET_MASK(uctxt->flags, DMA_RTAIL))
+		rcvctrl_ops |= HFI1_RCVCTRL_TAILUPD_ENB;
+	hfi1_rcvctrl(uctxt->dd, rcvctrl_ops, uctxt->ctxt);
+
+	/* Notify any waiting slaves */
+	if (uctxt->subctxt_cnt) {
+		clear_bit(HFI1_CTXT_MASTER_UNINIT, &uctxt->event_flags);
+		wake_up(&uctxt->wait);
+	}
+	ret = 0;
+
+done:
+	return ret;
+}
+
+static int get_ctxt_info(struct file *fp, void __user *ubase, __u32 len)
+{
+	struct hfi1_ctxt_info cinfo;
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_filedata *fd = fp->private_data;
+	int ret = 0;
+
+	ret = hfi1_get_base_kinfo(uctxt, &cinfo);
+	if (ret < 0)
+		goto done;
+	cinfo.num_active = hfi1_count_active_units();
+	cinfo.unit = uctxt->dd->unit;
+	cinfo.ctxt = uctxt->ctxt;
+	cinfo.subctxt = subctxt_fp(fp);
+	cinfo.rcvtids = roundup(uctxt->egrbufs.alloced,
+				uctxt->dd->rcv_entries.group_size) +
+		uctxt->expected_count;
+	cinfo.credits = uctxt->sc->credits;
+	cinfo.numa_node = uctxt->numa_id;
+	cinfo.rec_cpu = fd->rec_cpu_num;
+	cinfo.send_ctxt = uctxt->sc->hw_context;
+
+	cinfo.egrtids = uctxt->egrbufs.alloced;
+	cinfo.rcvhdrq_cnt = uctxt->rcvhdrq_cnt;
+	cinfo.rcvhdrq_entsize = uctxt->rcvhdrqentsize << 2;
+	cinfo.sdma_ring_size = user_sdma_comp_fp(fp)->nentries;
+	cinfo.rcvegr_size = uctxt->egrbufs.rcvtid_size;
+
+	trace_hfi1_ctxt_info(uctxt->dd, uctxt->ctxt, subctxt_fp(fp), cinfo);
+	if (copy_to_user(ubase, &cinfo, sizeof(cinfo)))
+		ret = -EFAULT;
+done:
+	return ret;
+}
+
+static int setup_ctxt(struct file *fp)
+{
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_devdata *dd = uctxt->dd;
+	int ret = 0;
+
+	/*
+	 * Context should be set up only once (including allocation and
+	 * programming of eager buffers. This is done if context sharing
+	 * is not requested or by the master process.
+	 */
+	if (!uctxt->subctxt_cnt || !subctxt_fp(fp)) {
+		ret = hfi1_init_ctxt(uctxt->sc);
+		if (ret)
+			goto done;
+
+		/* Now allocate the RcvHdr queue and eager buffers. */
+		ret = hfi1_create_rcvhdrq(dd, uctxt);
+		if (ret)
+			goto done;
+		ret = hfi1_setup_eagerbufs(uctxt);
+		if (ret)
+			goto done;
+		if (uctxt->subctxt_cnt && !subctxt_fp(fp)) {
+			ret = setup_subctxt(uctxt);
+			if (ret)
+				goto done;
+		}
+		/* Setup Expected Rcv memories */
+		uctxt->tid_pg_list = vzalloc(uctxt->expected_count *
+					     sizeof(struct page **));
+		if (!uctxt->tid_pg_list) {
+			ret = -ENOMEM;
+			goto done;
+		}
+		uctxt->physshadow = vzalloc(uctxt->expected_count *
+					    sizeof(*uctxt->physshadow));
+		if (!uctxt->physshadow) {
+			ret = -ENOMEM;
+			goto done;
+		}
+		/* allocate expected TID map and initialize the cursor */
+		atomic_set(&uctxt->tidcursor, 0);
+		uctxt->numtidgroups = uctxt->expected_count /
+			dd->rcv_entries.group_size;
+		uctxt->tidmapcnt = uctxt->numtidgroups / BITS_PER_LONG +
+			!!(uctxt->numtidgroups % BITS_PER_LONG);
+		uctxt->tidusemap = kzalloc_node(uctxt->tidmapcnt *
+						sizeof(*uctxt->tidusemap),
+						GFP_KERNEL, uctxt->numa_id);
+		if (!uctxt->tidusemap) {
+			ret = -ENOMEM;
+			goto done;
+		}
+		/*
+		 * In case that the number of groups is not a multiple of
+		 * 64 (the number of groups in a tidusemap element), mark
+		 * the extra ones as used. This will effectively make them
+		 * permanently used and should never be assigned. Otherwise,
+		 * the code which checks how many free groups we have will
+		 * get completely confused about the state of the bits.
+		 */
+		if (uctxt->numtidgroups % BITS_PER_LONG)
+			uctxt->tidusemap[uctxt->tidmapcnt - 1] =
+				~((1ULL << (uctxt->numtidgroups %
+					    BITS_PER_LONG)) - 1);
+		trace_hfi1_exp_tid_map(uctxt->ctxt, subctxt_fp(fp), 0,
+				       uctxt->tidusemap, uctxt->tidmapcnt);
+	}
+	ret = hfi1_user_sdma_alloc_queues(uctxt, fp);
+	if (ret)
+		goto done;
+
+	set_bit(HFI1_CTXT_SETUP_DONE, &uctxt->event_flags);
+done:
+	return ret;
+}
+
+static int get_base_info(struct file *fp, void __user *ubase, __u32 len)
+{
+	struct hfi1_base_info binfo;
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_devdata *dd = uctxt->dd;
+	ssize_t sz;
+	unsigned offset;
+	int ret = 0;
+
+	trace_hfi1_uctxtdata(uctxt->dd, uctxt);
+
+	memset(&binfo, 0, sizeof(binfo));
+	binfo.hw_version = dd->revision;
+	binfo.sw_version = HFI1_KERN_SWVERSION;
+	binfo.bthqp = kdeth_qp;
+	binfo.jkey = uctxt->jkey;
+	/*
+	 * If more than 64 contexts are enabled the allocated credit
+	 * return will span two or three contiguous pages. Since we only
+	 * map the page containing the context's credit return address,
+	 * we need to calculate the offset in the proper page.
+	 */
+	offset = ((u64)uctxt->sc->hw_free -
+		  (u64)dd->cr_base[uctxt->numa_id].va) % PAGE_SIZE;
+	binfo.sc_credits_addr = HFI1_MMAP_TOKEN(PIO_CRED, uctxt->ctxt,
+					       subctxt_fp(fp), offset);
+	binfo.pio_bufbase = HFI1_MMAP_TOKEN(PIO_BUFS, uctxt->ctxt,
+					    subctxt_fp(fp),
+					    uctxt->sc->base_addr);
+	binfo.pio_bufbase_sop = HFI1_MMAP_TOKEN(PIO_BUFS_SOP,
+						uctxt->ctxt,
+						subctxt_fp(fp),
+						uctxt->sc->base_addr);
+	binfo.rcvhdr_bufbase = HFI1_MMAP_TOKEN(RCV_HDRQ, uctxt->ctxt,
+					       subctxt_fp(fp),
+					       uctxt->rcvhdrq);
+	binfo.rcvegr_bufbase = HFI1_MMAP_TOKEN(RCV_EGRBUF, uctxt->ctxt,
+					       subctxt_fp(fp),
+					       uctxt->egrbufs.rcvtids[0].phys);
+	binfo.sdma_comp_bufbase = HFI1_MMAP_TOKEN(SDMA_COMP, uctxt->ctxt,
+						 subctxt_fp(fp), 0);
+	/*
+	 * user regs are at
+	 * (RXE_PER_CONTEXT_USER + (ctxt * RXE_PER_CONTEXT_SIZE))
+	 */
+	binfo.user_regbase = HFI1_MMAP_TOKEN(UREGS, uctxt->ctxt,
+					    subctxt_fp(fp), 0);
+	offset = ((((uctxt->ctxt - dd->first_user_ctxt) *
+		    HFI1_MAX_SHARED_CTXTS) + subctxt_fp(fp)) *
+		  sizeof(*dd->events)) & ~PAGE_MASK;
+	binfo.events_bufbase = HFI1_MMAP_TOKEN(EVENTS, uctxt->ctxt,
+					      subctxt_fp(fp),
+					      offset);
+	binfo.status_bufbase = HFI1_MMAP_TOKEN(STATUS, uctxt->ctxt,
+					      subctxt_fp(fp),
+					      dd->status);
+	if (HFI1_CAP_IS_USET(DMA_RTAIL))
+		binfo.rcvhdrtail_base = HFI1_MMAP_TOKEN(RTAIL, uctxt->ctxt,
+						       subctxt_fp(fp), 0);
+	if (uctxt->subctxt_cnt) {
+		binfo.subctxt_uregbase = HFI1_MMAP_TOKEN(SUBCTXT_UREGS,
+							uctxt->ctxt,
+							subctxt_fp(fp), 0);
+		binfo.subctxt_rcvhdrbuf = HFI1_MMAP_TOKEN(SUBCTXT_RCV_HDRQ,
+							 uctxt->ctxt,
+							 subctxt_fp(fp), 0);
+		binfo.subctxt_rcvegrbuf = HFI1_MMAP_TOKEN(SUBCTXT_EGRBUF,
+							 uctxt->ctxt,
+							 subctxt_fp(fp), 0);
+	}
+	sz = (len < sizeof(binfo)) ? len : sizeof(binfo);
+	if (copy_to_user(ubase, &binfo, sz))
+		ret = -EFAULT;
+	return ret;
+}
+
+static unsigned int poll_urgent(struct file *fp,
+				struct poll_table_struct *pt)
+{
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned pollflag;
+
+	poll_wait(fp, &uctxt->wait, pt);
+
+	spin_lock_irq(&dd->uctxt_lock);
+	if (uctxt->urgent != uctxt->urgent_poll) {
+		pollflag = POLLIN | POLLRDNORM;
+		uctxt->urgent_poll = uctxt->urgent;
+	} else {
+		pollflag = 0;
+		set_bit(HFI1_CTXT_WAITING_URG, &uctxt->event_flags);
+	}
+	spin_unlock_irq(&dd->uctxt_lock);
+
+	return pollflag;
+}
+
+static unsigned int poll_next(struct file *fp,
+			      struct poll_table_struct *pt)
+{
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned pollflag;
+
+	poll_wait(fp, &uctxt->wait, pt);
+
+	spin_lock_irq(&dd->uctxt_lock);
+	if (hdrqempty(uctxt)) {
+		set_bit(HFI1_CTXT_WAITING_RCV, &uctxt->event_flags);
+		hfi1_rcvctrl(dd, HFI1_RCVCTRL_INTRAVAIL_ENB, uctxt->ctxt);
+		pollflag = 0;
+	} else
+		pollflag = POLLIN | POLLRDNORM;
+	spin_unlock_irq(&dd->uctxt_lock);
+
+	return pollflag;
+}
+
+/*
+ * Find all user contexts in use, and set the specified bit in their
+ * event mask.
+ * See also find_ctxt() for a similar use, that is specific to send buffers.
+ */
+int hfi1_set_uevent_bits(struct hfi1_pportdata *ppd, const int evtbit)
+{
+	struct hfi1_ctxtdata *uctxt;
+	struct hfi1_devdata *dd = ppd->dd;
+	unsigned ctxt;
+	int ret = 0;
+	unsigned long flags;
+
+	if (!dd->events) {
+		ret = -EINVAL;
+		goto done;
+	}
+
+	spin_lock_irqsave(&dd->uctxt_lock, flags);
+	for (ctxt = dd->first_user_ctxt; ctxt < dd->num_rcv_contexts;
+	     ctxt++) {
+		uctxt = dd->rcd[ctxt];
+		if (uctxt) {
+			unsigned long *evs = dd->events +
+				(uctxt->ctxt - dd->first_user_ctxt) *
+				HFI1_MAX_SHARED_CTXTS;
+			int i;
+			/*
+			 * subctxt_cnt is 0 if not shared, so do base
+			 * separately, first, then remaining subctxt, if any
+			 */
+			set_bit(evtbit, evs);
+			for (i = 1; i < uctxt->subctxt_cnt; i++)
+				set_bit(evtbit, evs + i);
+		}
+	}
+	spin_unlock_irqrestore(&dd->uctxt_lock, flags);
+done:
+	return ret;
+}
+
+/**
+ * manage_rcvq - manage a context's receive queue
+ * @uctxt: the context
+ * @subctxt: the sub-context
+ * @start_stop: action to carry out
+ *
+ * start_stop == 0 disables receive on the context, for use in queue
+ * overflow conditions.  start_stop==1 re-enables, to be used to
+ * re-init the software copy of the head register
+ */
+static int manage_rcvq(struct hfi1_ctxtdata *uctxt, unsigned subctxt,
+		       int start_stop)
+{
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned int rcvctrl_op;
+
+	if (subctxt)
+		goto bail;
+	/* atomically clear receive enable ctxt. */
+	if (start_stop) {
+		/*
+		 * On enable, force in-memory copy of the tail register to
+		 * 0, so that protocol code doesn't have to worry about
+		 * whether or not the chip has yet updated the in-memory
+		 * copy or not on return from the system call. The chip
+		 * always resets it's tail register back to 0 on a
+		 * transition from disabled to enabled.
+		 */
+		if (uctxt->rcvhdrtail_kvaddr)
+			clear_rcvhdrtail(uctxt);
+		rcvctrl_op = HFI1_RCVCTRL_CTXT_ENB;
+	} else
+		rcvctrl_op = HFI1_RCVCTRL_CTXT_DIS;
+	hfi1_rcvctrl(dd, rcvctrl_op, uctxt->ctxt);
+	/* always; new head should be equal to new tail; see above */
+bail:
+	return 0;
+}
+
+/*
+ * clear the event notifier events for this context.
+ * User process then performs actions appropriate to bit having been
+ * set, if desired, and checks again in future.
+ */
+static int user_event_ack(struct hfi1_ctxtdata *uctxt, int subctxt,
+			  unsigned long events)
+{
+	int i;
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned long *evs;
+
+	if (!dd->events)
+		return 0;
+
+	evs = dd->events + ((uctxt->ctxt - dd->first_user_ctxt) *
+			    HFI1_MAX_SHARED_CTXTS) + subctxt;
+
+	for (i = 0; i <= _HFI1_MAX_EVENT_BIT; i++) {
+		if (!test_bit(i, &events))
+			continue;
+		clear_bit(i, evs);
+	}
+	return 0;
+}
+
+#define num_user_pages(vaddr, len)					\
+	(1 + (((((unsigned long)(vaddr) +				\
+		 (unsigned long)(len) - 1) & PAGE_MASK) -		\
+	       ((unsigned long)vaddr & PAGE_MASK)) >> PAGE_SHIFT))
+
+/**
+ * tzcnt - count the number of trailing zeros in a 64bit value
+ * @value: the value to be examined
+ *
+ * Returns the number of trailing least significant zeros in the
+ * the input value. If the value is zero, return the number of
+ * bits of the value.
+ */
+static inline u8 tzcnt(u64 value)
+{
+	return value ? __builtin_ctzl(value) : sizeof(value) * 8;
+}
+
+static inline unsigned num_free_groups(unsigned long map, u16 *start)
+{
+	unsigned free;
+	u16 bitidx = *start;
+
+	if (bitidx >= BITS_PER_LONG)
+		return 0;
+	/* "Turn off" any bits set before our bit index */
+	map &= ~((1ULL << bitidx) - 1);
+	free = tzcnt(map) - bitidx;
+	while (!free && bitidx < BITS_PER_LONG) {
+		/* Zero out the last set bit so we look at the rest */
+		map &= ~(1ULL << bitidx);
+		/*
+		 * Account for the previously checked bits and advance
+		 * the bit index. We don't have to check for bitidx
+		 * getting bigger than BITS_PER_LONG here as it would
+		 * mean extra instructions that we don't need. If it
+		 * did happen, it would push free to a negative value
+		 * which will break the loop.
+		 */
+		free = tzcnt(map) - ++bitidx;
+	}
+	*start = bitidx;
+	return free;
+}
+
+static int exp_tid_setup(struct file *fp, struct hfi1_tid_info *tinfo)
+{
+	int ret = 0;
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned tid, mapped = 0, npages, ngroups, exp_groups,
+		tidpairs = uctxt->expected_count / 2;
+	struct page **pages;
+	unsigned long vaddr, tidmap[uctxt->tidmapcnt];
+	dma_addr_t *phys;
+	u32 tidlist[tidpairs], pairidx = 0, tidcursor;
+	u16 useidx, idx, bitidx, tidcnt = 0;
+
+	vaddr = tinfo->vaddr;
+
+	if (vaddr & ~PAGE_MASK) {
+		ret = -EINVAL;
+		goto bail;
+	}
+
+	npages = num_user_pages(vaddr, tinfo->length);
+	if (!npages) {
+		ret = -EINVAL;
+		goto bail;
+	}
+	if (!access_ok(VERIFY_WRITE, (void __user *)vaddr,
+		       npages * PAGE_SIZE)) {
+		dd_dev_err(dd, "Fail vaddr %p, %u pages, !access_ok\n",
+			   (void *)vaddr, npages);
+		ret = -EFAULT;
+		goto bail;
+	}
+
+	memset(tidmap, 0, sizeof(tidmap[0]) * uctxt->tidmapcnt);
+	memset(tidlist, 0, sizeof(tidlist[0]) * tidpairs);
+
+	exp_groups = uctxt->expected_count / dd->rcv_entries.group_size;
+	/* which group set do we look at first? */
+	tidcursor = atomic_read(&uctxt->tidcursor);
+	useidx = (tidcursor >> 16) & 0xffff;
+	bitidx = tidcursor & 0xffff;
+
+	/*
+	 * Keep going until we've mapped all pages or we've exhausted all
+	 * RcvArray entries.
+	 * This iterates over the number of tidmaps + 1
+	 * (idx <= uctxt->tidmapcnt) so we check the bitmap which we
+	 * started from one more time for any free bits before the
+	 * starting point bit.
+	 */
+	for (mapped = 0, idx = 0;
+	     mapped < npages && idx <= uctxt->tidmapcnt;) {
+		u64 i, offset = 0;
+		unsigned free, pinned, pmapped = 0, bits_used;
+		u16 grp;
+
+		/*
+		 * "Reserve" the needed group bits under lock so other
+		 * processes can't step in the middle of it. Once
+		 * reserved, we don't need the lock anymore since we
+		 * are guaranteed the groups.
+		 */
+		spin_lock(&uctxt->exp_lock);
+		if (uctxt->tidusemap[useidx] == -1ULL ||
+		    bitidx >= BITS_PER_LONG) {
+			/* no free groups in the set, use the next */
+			useidx = (useidx + 1) % uctxt->tidmapcnt;
+			idx++;
+			bitidx = 0;
+			spin_unlock(&uctxt->exp_lock);
+			continue;
+		}
+		ngroups = ((npages - mapped) / dd->rcv_entries.group_size) +
+			!!((npages - mapped) % dd->rcv_entries.group_size);
+
+		/*
+		 * If we've gotten here, the current set of groups does have
+		 * one or more free groups.
+		 */
+		free = num_free_groups(uctxt->tidusemap[useidx], &bitidx);
+		if (!free) {
+			/*
+			 * Despite the check above, free could still come back
+			 * as 0 because we don't check the entire bitmap but
+			 * we start from bitidx.
+			 */
+			spin_unlock(&uctxt->exp_lock);
+			continue;
+		}
+		bits_used = min(free, ngroups);
+		tidmap[useidx] |= ((1ULL << bits_used) - 1) << bitidx;
+		uctxt->tidusemap[useidx] |= tidmap[useidx];
+		spin_unlock(&uctxt->exp_lock);
+
+		/*
+		 * At this point, we know where in the map we have free bits.
+		 * properly offset into the various "shadow" arrays and compute
+		 * the RcvArray entry index.
+		 */
+		offset = ((useidx * BITS_PER_LONG) + bitidx) *
+			dd->rcv_entries.group_size;
+		pages = uctxt->tid_pg_list + offset;
+		phys = uctxt->physshadow + offset;
+		tid = uctxt->expected_base + offset;
+
+		/* Calculate how many pages we can pin based on free bits */
+		pinned = min((bits_used * dd->rcv_entries.group_size),
+			     (npages - mapped));
+		/*
+		 * Now that we know how many free RcvArray entries we have,
+		 * we can pin that many user pages.
+		 */
+		ret = hfi1_get_user_pages(vaddr + (mapped * PAGE_SIZE),
+					  pinned, pages);
+		if (ret) {
+			/*
+			 * We can't continue because the pages array won't be
+			 * initialized. This should never happen,
+			 * unless perhaps the user has mpin'ed the pages
+			 * themselves.
+			 */
+			dd_dev_info(dd,
+				    "Failed to lock addr %p, %u pages: errno %d\n",
+				    (void *) vaddr, pinned, -ret);
+			/*
+			 * Let go of the bits that we reserved since we are not
+			 * going to use them.
+			 */
+			spin_lock(&uctxt->exp_lock);
+			uctxt->tidusemap[useidx] &=
+				~(((1ULL << bits_used) - 1) << bitidx);
+			spin_unlock(&uctxt->exp_lock);
+			goto done;
+		}
+		/*
+		 * How many groups do we need based on how many pages we have
+		 * pinned?
+		 */
+		ngroups = (pinned / dd->rcv_entries.group_size) +
+			!!(pinned % dd->rcv_entries.group_size);
+		/*
+		 * Keep programming RcvArray entries for all the <ngroups> free
+		 * groups.
+		 */
+		for (i = 0, grp = 0; grp < ngroups; i++, grp++) {
+			unsigned j;
+			u32 pair_size = 0, tidsize;
+			/*
+			 * This inner loop will program an entire group or the
+			 * array of pinned pages (which ever limit is hit
+			 * first).
+			 */
+			for (j = 0; j < dd->rcv_entries.group_size &&
+				     pmapped < pinned; j++, pmapped++, tid++) {
+				tidsize = PAGE_SIZE;
+				phys[pmapped] = hfi1_map_page(dd->pcidev,
+						   pages[pmapped], 0,
+						   tidsize, PCI_DMA_FROMDEVICE);
+				trace_hfi1_exp_rcv_set(uctxt->ctxt,
+						       subctxt_fp(fp),
+						       tid, vaddr,
+						       phys[pmapped],
+						       pages[pmapped]);
+				/*
+				 * Each RcvArray entry is programmed with one
+				 * page * worth of memory. This will handle
+				 * the 8K MTU as well as anything smaller
+				 * due to the fact that both entries in the
+				 * RcvTidPair are programmed with a page.
+				 * PSM currently does not handle anything
+				 * bigger than 8K MTU, so should we even worry
+				 * about 10K here?
+				 */
+				hfi1_put_tid(dd, tid, PT_EXPECTED,
+					     phys[pmapped],
+					     ilog2(tidsize >> PAGE_SHIFT) + 1);
+				pair_size += tidsize >> PAGE_SHIFT;
+				EXP_TID_RESET(tidlist[pairidx], LEN, pair_size);
+				if (!(tid % 2)) {
+					tidlist[pairidx] |=
+					   EXP_TID_SET(IDX,
+						(tid - uctxt->expected_base)
+						       / 2);
+					tidlist[pairidx] |=
+						EXP_TID_SET(CTRL, 1);
+					tidcnt++;
+				} else {
+					tidlist[pairidx] |=
+						EXP_TID_SET(CTRL, 2);
+					pair_size = 0;
+					pairidx++;
+				}
+			}
+			/*
+			 * We've programmed the entire group (or as much of the
+			 * group as we'll use. Now, it's time to push it out...
+			 */
+			flush_wc();
+		}
+		mapped += pinned;
+		atomic_set(&uctxt->tidcursor,
+			   (((useidx & 0xffffff) << 16) |
+			    ((bitidx + bits_used) & 0xffffff)));
+	}
+	trace_hfi1_exp_tid_map(uctxt->ctxt, subctxt_fp(fp), 0, uctxt->tidusemap,
+			       uctxt->tidmapcnt);
+
+done:
+	/* If we've mapped anything, copy relevant info to user */
+	if (mapped) {
+		if (copy_to_user((void __user *)(unsigned long)tinfo->tidlist,
+				 tidlist, sizeof(tidlist[0]) * tidcnt)) {
+			ret = -EFAULT;
+			goto done;
+		}
+		/* copy TID info to user */
+		if (copy_to_user((void __user *)(unsigned long)tinfo->tidmap,
+				 tidmap, sizeof(tidmap[0]) * uctxt->tidmapcnt))
+			ret = -EFAULT;
+	}
+bail:
+	/*
+	 * Calculate mapped length. New Exp TID protocol does not "unwind" and
+	 * report an error if it can't map the entire buffer. It just reports
+	 * the length that was mapped.
+	 */
+	tinfo->length = mapped * PAGE_SIZE;
+	tinfo->tidcnt = tidcnt;
+	return ret;
+}
+
+static int exp_tid_free(struct file *fp, struct hfi1_tid_info *tinfo)
+{
+	struct hfi1_ctxtdata *uctxt = ctxt_fp(fp);
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned long tidmap[uctxt->tidmapcnt];
+	struct page **pages;
+	dma_addr_t *phys;
+	u16 idx, bitidx, tid;
+	int ret = 0;
+
+	if (copy_from_user(&tidmap, (void __user *)(unsigned long)
+			   tinfo->tidmap,
+			   sizeof(tidmap[0]) * uctxt->tidmapcnt)) {
+		ret = -EFAULT;
+		goto done;
+	}
+	for (idx = 0; idx < uctxt->tidmapcnt; idx++) {
+		unsigned long map;
+
+		bitidx = 0;
+		if (!tidmap[idx])
+			continue;
+		map = tidmap[idx];
+		while ((bitidx = tzcnt(map)) < BITS_PER_LONG) {
+			int i, pcount = 0;
+			struct page *pshadow[dd->rcv_entries.group_size];
+			unsigned offset = ((idx * BITS_PER_LONG) + bitidx) *
+				dd->rcv_entries.group_size;
+
+			pages = uctxt->tid_pg_list + offset;
+			phys = uctxt->physshadow + offset;
+			tid = uctxt->expected_base + offset;
+			for (i = 0; i < dd->rcv_entries.group_size;
+			     i++, tid++) {
+				if (pages[i]) {
+					hfi1_put_tid(dd, tid, PT_INVALID,
+						      0, 0);
+					trace_hfi1_exp_rcv_free(uctxt->ctxt,
+								subctxt_fp(fp),
+								tid, phys[i],
+								pages[i]);
+					pci_unmap_page(dd->pcidev, phys[i],
+					      PAGE_SIZE, PCI_DMA_FROMDEVICE);
+					pshadow[pcount] = pages[i];
+					pages[i] = NULL;
+					pcount++;
+					phys[i] = 0;
+				}
+			}
+			flush_wc();
+			hfi1_release_user_pages(pshadow, pcount);
+			clear_bit(bitidx, &uctxt->tidusemap[idx]);
+			map &= ~(1ULL<<bitidx);
+		}
+	}
+	trace_hfi1_exp_tid_map(uctxt->ctxt, subctxt_fp(fp), 1, uctxt->tidusemap,
+			       uctxt->tidmapcnt);
+done:
+	return ret;
+}
+
+static void unlock_exp_tids(struct hfi1_ctxtdata *uctxt)
+{
+	struct hfi1_devdata *dd = uctxt->dd;
+	unsigned tid;
+
+	dd_dev_info(dd, "ctxt %u unlocking any locked expTID pages\n",
+		    uctxt->ctxt);
+	for (tid = 0; tid < uctxt->expected_count; tid++) {
+		struct page *p = uctxt->tid_pg_list[tid];
+		dma_addr_t phys;
+
+		if (!p)
+			continue;
+
+		phys = uctxt->physshadow[tid];
+		uctxt->physshadow[tid] = 0;
+		uctxt->tid_pg_list[tid] = NULL;
+		pci_unmap_page(dd->pcidev, phys, PAGE_SIZE, PCI_DMA_FROMDEVICE);
+		hfi1_release_user_pages(&p, 1);
+	}
+}
+
+static int set_ctxt_pkey(struct hfi1_ctxtdata *uctxt, unsigned subctxt,
+			 u16 pkey)
+{
+	int ret = -ENOENT, i, intable = 0;
+	struct hfi1_pportdata *ppd = uctxt->ppd;
+	struct hfi1_devdata *dd = uctxt->dd;
+
+	if (pkey == LIM_MGMT_P_KEY || pkey == FULL_MGMT_P_KEY) {
+		ret = -EINVAL;
+		goto done;
+	}
+
+	for (i = 0; i < ARRAY_SIZE(ppd->pkeys); i++)
+		if (pkey == ppd->pkeys[i]) {
+			intable = 1;
+			break;
+		}
+
+	if (intable)
+		ret = hfi1_set_ctxt_pkey(dd, uctxt->ctxt, pkey);
+done:
+	return ret;
+}
+
+static int ui_open(struct inode *inode, struct file *filp)
+{
+	struct hfi1_devdata *dd;
+
+	dd = container_of(inode->i_cdev, struct hfi1_devdata, ui_cdev);
+	filp->private_data = dd; /* for other methods */
+	return 0;
+}
+
+static int ui_release(struct inode *inode, struct file *filp)
+{
+	/* nothing to do */
+	return 0;
+}
+
+static loff_t ui_lseek(struct file *filp, loff_t offset, int whence)
+{
+	struct hfi1_devdata *dd = filp->private_data;
+
+	switch (whence) {
+	case SEEK_SET:
+		break;
+	case SEEK_CUR:
+		offset += filp->f_pos;
+		break;
+	case SEEK_END:
+		offset = ((dd->kregend - dd->kregbase) + DC8051_DATA_MEM_SIZE) -
+			offset;
+		break;
+	default:
+		return -EINVAL;
+	}
+
+	if (offset < 0)
+		return -EINVAL;
+
+	if (offset >= (dd->kregend - dd->kregbase) + DC8051_DATA_MEM_SIZE)
+		return -EINVAL;
+
+	filp->f_pos = offset;
+
+	return filp->f_pos;
+}
+
+
+/* NOTE: assumes unsigned long is 8 bytes */
+static ssize_t ui_read(struct file *filp, char __user *buf, size_t count,
+			loff_t *f_pos)
+{
+	struct hfi1_devdata *dd = filp->private_data;
+	void __iomem *base = dd->kregbase;
+	unsigned long total, csr_off,
+		barlen = (dd->kregend - dd->kregbase);
+	u64 data;
+
+	/* only read 8 byte quantities */
+	if ((count % 8) != 0)
+		return -EINVAL;
+	/* offset must be 8-byte aligned */
+	if ((*f_pos % 8) != 0)
+		return -EINVAL;
+	/* destination buffer must be 8-byte aligned */
+	if ((unsigned long)buf % 8 != 0)
+		return -EINVAL;
+	/* must be in range */
+	if (*f_pos + count > (barlen + DC8051_DATA_MEM_SIZE))
+		return -EINVAL;
+	/* only set the base if we are not starting past the BAR */
+	if (*f_pos < barlen)
+		base += *f_pos;
+	csr_off = *f_pos;
+	for (total = 0; total < count; total += 8, csr_off += 8) {
+		/* accessing LCB CSRs requires more checks */
+		if (is_lcb_offset(csr_off)) {
+			if (read_lcb_csr(dd, csr_off, (u64 *)&data))
+				break; /* failed */
+		}
+		/*
+		 * Cannot read ASIC GPIO/QSFP* clear and force CSRs without a
+		 * false parity error.  Avoid the whole issue by not reading
+		 * them.  These registers are defined as having a read value
+		 * of 0.
+		 */
+		else if (csr_off == ASIC_GPIO_CLEAR
+				|| csr_off == ASIC_GPIO_FORCE
+				|| csr_off == ASIC_QSFP1_CLEAR
+				|| csr_off == ASIC_QSFP1_FORCE
+				|| csr_off == ASIC_QSFP2_CLEAR
+				|| csr_off == ASIC_QSFP2_FORCE)
+			data = 0;
+		else if (csr_off >= barlen) {
+			/*
+			 * read_8051_data can read more than just 8 bytes at
+			 * a time. However, folding this into the loop and
+			 * handling the reads in 8 byte increments allows us
+			 * to smoothly transition from chip memory to 8051
+			 * memory.
+			 */
+			if (read_8051_data(dd,
+					   (u32)(csr_off - barlen),
+					   sizeof(data), &data))
+				break; /* failed */
+		} else
+			data = readq(base + total);
+		if (put_user(data, (unsigned long __user *)(buf + total)))
+			break;
+	}
+	*f_pos += total;
+	return total;
+}
+
+/* NOTE: assumes unsigned long is 8 bytes */
+static ssize_t ui_write(struct file *filp, const char __user *buf,
+			size_t count, loff_t *f_pos)
+{
+	struct hfi1_devdata *dd = filp->private_data;
+	void __iomem *base;
+	unsigned long total, data, csr_off;
+	int in_lcb;
+
+	/* only write 8 byte quantities */
+	if ((count % 8) != 0)
+		return -EINVAL;
+	/* offset must be 8-byte aligned */
+	if ((*f_pos % 8) != 0)
+		return -EINVAL;
+	/* source buffer must be 8-byte aligned */
+	if ((unsigned long)buf % 8 != 0)
+		return -EINVAL;
+	/* must be in range */
+	if (*f_pos + count > dd->kregend - dd->kregbase)
+		return -EINVAL;
+
+	base = (void __iomem *)dd->kregbase + *f_pos;
+	csr_off = *f_pos;
+	in_lcb = 0;
+	for (total = 0; total < count; total += 8, csr_off += 8) {
+		if (get_user(data, (unsigned long __user *)(buf + total)))
+			break;
+		/* accessing LCB CSRs requires a special procedure */
+		if (is_lcb_offset(csr_off)) {
+			if (!in_lcb) {
+				int ret = acquire_lcb_access(dd, 1);
+
+				if (ret)
+					break;
+				in_lcb = 1;
+			}
+		} else {
+			if (in_lcb) {
+				release_lcb_access(dd, 1);
+				in_lcb = 0;
+			}
+		}
+		writeq(data, base + total);
+	}
+	if (in_lcb)
+		release_lcb_access(dd, 1);
+	*f_pos += total;
+	return total;
+}
+
+static const struct file_operations ui_file_ops = {
+	.owner = THIS_MODULE,
+	.llseek = ui_lseek,
+	.read = ui_read,
+	.write = ui_write,
+	.open = ui_open,
+	.release = ui_release,
+};
+#define UI_OFFSET 192	/* device minor offset for UI devices */
+static int create_ui = 1;
+
+static struct cdev wildcard_cdev;
+static struct device *wildcard_device;
+
+static atomic_t user_count = ATOMIC_INIT(0);
+
+static void user_remove(struct hfi1_devdata *dd)
+{
+	if (atomic_dec_return(&user_count) == 0)
+		hfi1_cdev_cleanup(&wildcard_cdev, &wildcard_device);
+
+	hfi1_cdev_cleanup(&dd->user_cdev, &dd->user_device);
+	hfi1_cdev_cleanup(&dd->ui_cdev, &dd->ui_device);
+}
+
+static int user_add(struct hfi1_devdata *dd)
+{
+	char name[10];
+	int ret;
+
+	if (atomic_inc_return(&user_count) == 1) {
+		ret = hfi1_cdev_init(0, class_name(), &hfi1_file_ops,
+				     &wildcard_cdev, &wildcard_device);
+		if (ret)
+			goto done;
+	}
+
+	snprintf(name, sizeof(name), "%s_%d", class_name(), dd->unit);
+	ret = hfi1_cdev_init(dd->unit + 1, name, &hfi1_file_ops,
+			     &dd->user_cdev, &dd->user_device);
+	if (ret)
+		goto done;
+
+	if (create_ui) {
+		snprintf(name, sizeof(name),
+			 "%s_ui%d", class_name(), dd->unit);
+		ret = hfi1_cdev_init(dd->unit + UI_OFFSET, name, &ui_file_ops,
+				     &dd->ui_cdev, &dd->ui_device);
+		if (ret)
+			goto done;
+	}
+
+	return 0;
+done:
+	user_remove(dd);
+	return ret;
+}
+
+/*
+ * Create per-unit files in /dev
+ */
+int hfi1_device_create(struct hfi1_devdata *dd)
+{
+	int r, ret;
+
+	r = user_add(dd);
+	ret = hfi1_diag_add(dd);
+	if (r && !ret)
+		ret = r;
+	return ret;
+}
+
+/*
+ * Remove per-unit files in /dev
+ * void, core kernel returns no errors for this stuff
+ */
+void hfi1_device_remove(struct hfi1_devdata *dd)
+{
+	user_remove(dd);
+	hfi1_diag_remove(dd);
+}