On 2018-09-27 11:12 AM, Keith Busch wrote:
> Reviewed-by: Keith Busch
Thanks for the reviews Keith!
Logan
and cleared on any error.
It also seems to be unnecessary to accumulate the length as the map_sgl
functions should only ever be called once per request.
Signed-off-by: Logan Gunthorpe
Acked-by: Sagi Grimberg
Cc: Christoph Hellwig
---
drivers/nvme/target/core.c | 18 ++
drivers
.
Signed-off-by: Logan Gunthorpe
Acked-by: Bjorn Helgaas
---
drivers/pci/p2pdma.c | 43 ++
include/linux/memremap.h | 1 +
include/linux/pci-p2pdma.h | 7 +++
3 files changed, 51 insertions(+)
diff --git a/drivers/pci/p2pdma.c b/drivers/pci
is P2P the entire SGL should be P2P.
Signed-off-by: Logan Gunthorpe
Reviewed-by: Christoph Hellwig
Reviewed-by: Sagi Grimberg
---
drivers/infiniband/core/rw.c | 11 +--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/core/rw.c b/drivers/infiniband/core
Add a sysfs group to display statistics about P2P memory that is
registered in each PCI device.
Attributes in the group display the total amount of P2P memory, the
amount available and whether it is published or not.
Signed-off-by: Logan Gunthorpe
Acked-by: Bjorn Helgaas
---
Documentation/ABI
QUEUE_FLAG_PCI_P2P is introduced meaning a driver's request queue
supports targeting P2P memory. This will be used by P2P providers and
orchestrators (in subsequent patches) to ensure block devices can
support P2P memory before submitting P2P backed pages to submit_bio().
Signed-off-by: Logan
converted to restructured text
at this time.
Signed-off-by: Logan Gunthorpe
Acked-by: Bjorn Helgaas
Cc: Jonathan Corbet
---
Documentation/driver-api/pci/index.rst | 1 +
Documentation/driver-api/pci/p2pdma.rst | 170
2 files changed, 171 insertions(+)
create mode
code]
Signed-off-by: Christoph Hellwig
Signed-off-by: Logan Gunthorpe
---
drivers/nvme/target/configfs.c| 36
drivers/nvme/target/core.c| 138 +-
drivers/nvme/target/io-cmd-bdev.c | 3 +
drivers/nvme/target/nvmet.h | 13 +++
drivers/nvme
Introduce a quirk to use CMB-like memory on older devices that have
an exposed BAR but do not advertise support for using CMBLOC and
CMBSIZE.
We'd like to use some of these older cards to test P2P memory.
Signed-off-by: Logan Gunthorpe
Reviewed-by: Sagi Grimberg
---
drivers/nvme/host/nvme.h
Add a new directory in the driver API guide for PCI specific
documentation.
This is in preparation for adding a new PCI P2P DMA driver writers
guide which will go in this directory.
Signed-off-by: Logan Gunthorpe
Cc: Jonathan Corbet
Cc: Mauro Carvalho Chehab
Cc: Greg Kroah-Hartman
Cc: Vinod
for attributes which take a boolean or a PCI device.
Any boolean as accepted by strtobool() turn P2P on or off (such as 'y', 'n',
'1', '0', etc). Specifying a full PCI device name/BDF will select the
specific device.
Signed-off-by: Logan Gunthorpe
Acked-by: Bjorn Helgaas
---
drivers/pci/p2pdma.c
spin) using switches from both Microsemi
and Broadcomm.
--
Logan Gunthorpe (13):
PCI/P2PDMA: Support peer-to-peer memory
PCI/P2PDMA: Add sysfs group to display p2pmem stats
PCI/P2PDMA: Add PCI p2pmem DMA mappings to adjust the bus offset
PCI/P2PDMA: Introduce configfs/sysfs enable attribute
capability bit to advertise whether this is possible for
future hardware.
This commit includes significant rework and feedback from Christoph
Hellwig.
Signed-off-by: Christoph Hellwig
Signed-off-by: Logan Gunthorpe
Acked-by: Bjorn Helgaas # PCI pieces
---
drivers/pci/Kconfig| 17
will not be supported by memremap()
and therefore will not be support PCI P2P and have no support for CMB.
Signed-off-by: Logan Gunthorpe
---
drivers/nvme/host/pci.c | 80 +++--
1 file changed, 45 insertions(+), 35 deletions(-)
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme
-by: Logan Gunthorpe
Reviewed-by: Sagi Grimberg
Reviewed-by: Christoph Hellwig
---
drivers/nvme/host/core.c | 4
drivers/nvme/host/nvme.h | 1 +
drivers/nvme/host/pci.c | 17 +
3 files changed, 18 insertions(+), 4 deletions(-)
diff --git a/drivers/nvme/host/core.c b/drivers
Thanks for the review Randy! I'll make the changes for the next time we
post the series.
On 22/05/18 03:24 PM, Randy Dunlap wrote:
>> +The first task an orchestrator driver must do is compile a list of
>> +all client drivers that will be involved in a given transaction. For
>> +example, the NVMe
On 5/11/2018 4:24 PM, Stephen Bates wrote:
All
Alex (or anyone else) can you point to where IOVA addresses are generated?
A case of RTFM perhaps (though a pointer to the code would still be
appreciated).
https://www.kernel.org/doc/Documentation/Intel-IOMMU.txt
Some exceptions to IOVA
On 5/11/2018 2:52 AM, Christian König wrote:
This only works when the IOVA and the PCI bus addresses never overlap.
I'm not sure how the IOVA allocation works but I don't think we
guarantee that on Linux.
I find this hard to believe. There's always the possibility that some
part of the
On 10/05/18 12:41 PM, Stephen Bates wrote:
> Hi Jerome
>
>>Note on GPU we do would not rely on ATS for peer to peer. Some part
>>of the GPU (DMA engines) do not necessarily support ATS. Yet those
>>are the part likely to be use in peer to peer.
>
> OK this is good to know. I agree
On 10/05/18 11:11 AM, Stephen Bates wrote:
>> Not to me. In the p2pdma code we specifically program DMA engines with
>> the PCI bus address.
>
> Ah yes of course. Brain fart on my part. We are not programming the P2PDMA
> initiator with an IOVA but with the PCI bus address...
>
>> So
On 10/05/18 08:16 AM, Stephen Bates wrote:
> Hi Christian
>
>> Why would a switch not identify that as a peer address? We use the PASID
>>together with ATS to identify the address space which a transaction
>>should use.
>
> I think you are conflating two types of TLPs here. If the
On 09/05/18 07:40 AM, Christian König wrote:
> The key takeaway is that when any device has ATS enabled you can't
> disable ACS without breaking it (even if you unplug and replug it).
I don't follow how you came to this conclusion...
The ACS bits we'd be turning off are the ones that force
On 08/05/18 05:11 PM, Alex Williamson wrote:
> On to the implementation details... I already mentioned the BDF issue
> in my other reply. If we had a way to persistently identify a device,
> would we specify the downstream points at which we want to disable ACS
> or the endpoints that we want
On 08/05/18 05:00 PM, Dan Williams wrote:
>> I'd advise caution with a user supplied BDF approach, we have no
>> guaranteed persistence for a device's PCI address. Adding a device
>> might renumber the buses, replacing a device with one that consumes
>> more/less bus numbers can renumber the
On 08/05/18 04:03 PM, Alex Williamson wrote:
> If IOMMU grouping implies device assignment (because nobody else uses
> it to the same extent as device assignment) then the build-time option
> falls to pieces, we need a single kernel that can do both. I think we
> need to get more clever about
On 08/05/18 02:43 PM, Alex Williamson wrote:
> Yes, GPUs seem to be leading the pack in implementing ATS. So now the
> dumb question, why not simply turn off the IOMMU and thus ACS? The
> argument of using the IOMMU for security is rather diminished if we're
> specifically enabling devices to
On 08/05/18 02:13 PM, Alex Williamson wrote:
> Well, I'm a bit confused, this patch series is specifically disabling
> ACS on switches, but per the spec downstream switch ports implementing
> ACS MUST implement direct translated P2P. So it seems the only
> potential gap here is the endpoint,
On 08/05/18 01:34 PM, Alex Williamson wrote:
> They are not so unrelated, see the ACS Direct Translated P2P
> capability, which in fact must be implemented by switch downstream
> ports implementing ACS and works specifically with ATS. This appears to
> be the way the PCI SIG would intend for
On 08/05/18 10:57 AM, Alex Williamson wrote:
> AIUI from previously questioning this, the change is hidden behind a
> build-time config option and only custom kernels or distros optimized
> for this sort of support would enable that build option. I'm more than
> a little dubious though that
On 08/05/18 10:50 AM, Christian König wrote:
> E.g. transactions are initially send to the root complex for
> translation, that's for sure. But at least for AMD GPUs the root complex
> answers with the translated address which is then cached in the device.
>
> So further transactions for the
On 08/05/18 01:17 AM, Christian König wrote:
> AMD APUs mandatory need the ACS flag set for the GPU integrated in the
> CPU when IOMMU is enabled or otherwise you will break SVM.
Well, given that the current set only disables ACS bits on bridges
(previous versions were only on switches) this
> How do you envison merging this? There's a big chunk in drivers/pci, but
> really no opportunity for conflicts there, and there's significant stuff in
> block and nvme that I don't really want to merge.
>
> If Alex is OK with the ACS situation, I can ack the PCI parts and you could
> merge it
Thanks for the review. I'll apply all of these for the changes for next
version of the set.
>> +/*
>> + * If a device is behind a switch, we try to find the upstream bridge
>> + * port of the switch. This requires two calls to pci_upstream_bridge():
>> + * one for the upstream port on the switch,
On 04/05/18 08:27 AM, Christian König wrote:
> Are you sure that this is more convenient? At least on first glance it
> feels overly complicated.
>
> I mean what's the difference between the two approaches?
>
> sum = pci_p2pdma_distance(target, [A, B, C, target]);
>
> and
>
> sum
On 03/05/18 11:29 AM, Christian König wrote:
> Ok, that is the point where I'm stuck. Why do we need that in one
> function call in the PCIe subsystem?
>
> The problem at least with GPUs is that we seriously don't have that
> information here, cause the PCI subsystem might not be aware of all
On 03/05/18 03:05 AM, Christian König wrote:
> Ok, I'm still missing the big picture here. First question is what is
> the P2PDMA provider?
Well there's some pretty good documentation in the patchset for this,
but in short, a provider is a device that provides some kind of P2P
resource (ie.
Hi Christian,
On 5/2/2018 5:51 AM, Christian König wrote:
it would be rather nice to have if you could separate out the functions
to detect if peer2peer is possible between two devices.
This would essentially be pci_p2pdma_distance() in the existing
patchset. It returns the sum of the
. The PCI-SIG may be exploring adding a new
capability bit to advertise whether this is possible for future
hardware.
This commit includes significant rework and feedback from Christoph
Hellwig.
Signed-off-by: Christoph Hellwig <h...@lst.de>
Signed-off-by: Logan Gunthorpe <log...@del
The DMA address used when mapping PCI P2P memory must be the PCI bus
address. Thus, introduce pci_p2pmem_[un]map_sg() to map the correct
addresses when using P2P memory.
For this, we assume that an SGL passed to these functions contain all
P2P memory or no P2P memory.
Signed-off-by: Logan
-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Sagi Grimberg <s...@grimberg.me>
Reviewed-by: Christoph Hellwig <h...@lst.de>
---
drivers/nvme/host/core.c | 4
drivers/nvme/host/nvme.h | 1 +
drivers/nvme/host/pci.c | 19 +++
3 files changed, 2
drivers. The presently
unused 'sq' argument in the alloc function will be necessary to
decide whether to use peer-to-peer memory and obtain the correct
provider to allocate the memory.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Cc: Christoph Hellwig <h...@lst.de>
Cc: Sagi
Add a sysfs group to display statistics about P2P memory that is
registered in each PCI device.
Attributes in the group display the total amount of P2P memory, the
amount available and whether it is published or not.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
Documentati
flag set.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Sagi Grimberg <s...@grimberg.me>
Reviewed-by: Christoph Hellwig <h...@lst.de>
---
block/blk-core.c | 3 +++
include/linux/blk_types.h | 18 +-
include/linux/blkdev.h| 3 ++
converted to restructured text
at this time.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Cc: Jonathan Corbet <cor...@lwn.net>
---
Documentation/PCI/index.rst | 14 +++
Documentation/driver-api/pci/index.rst | 1 +
Documentation/driver-api/pci/p2pd
be called once.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Cc: Christoph Hellwig <h...@lst.de>
Cc: Sagi Grimberg <s...@grimberg.me>
---
drivers/nvme/target/rdma.c | 20
1 file changed, 12 insertions(+), 8 deletions(-)
diff --git a/drivers/nvme/target/rdm
se <sw...@opengridcomputing.com>
[hch: partial rewrite of the initial code]
Signed-off-by: Christoph Hellwig <h...@lst.de>
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/nvme/target/configfs.c | 67 ++
drivers/nvme
transactions.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/pci/Kconfig| 9 +
drivers/pci/p2pdma.c | 45 ++---
drivers/pci/pci.c | 6 ++
include/linux/pci-p2pdma.h | 5 +
4 files chang
is P2P the entire SGL should be P2P.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Christoph Hellwig <h...@lst.de>
---
drivers/infiniband/core/rw.c | 13 +++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/core/rw
, devm_memremap_pages() allocates regular memory without
side effects that's accessible without the iomem accessors.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/nvme/host/pci.c | 75 +++--
1 file changed, 41 insertions(+), 34 del
Add a new directory in the driver API guide for PCI specific
documentation.
This is in preparation for adding a new PCI P2P DMA driver writers
guide which will go in this directory.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Cc: Jonathan Corbet <cor...@lwn.net>
Cc: Ma
Introduce a quirk to use CMB-like memory on older devices that have
an exposed BAR but do not advertise support for using CMBLOC and
CMBSIZE.
We'd like to use some of these older cards to test P2P memory.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Sagi Grimb
On 27/03/18 02:47 AM, Jonathan Cameron wrote:
> I'll see if I can get our PCI SIG people to follow this through and see if
> it is just an omission or as Bjorn suggested, there is some reason we
> aren't thinking of that makes it hard.
That would be great! Thanks!
Logan
On 26/03/18 01:35 PM, Jason Gunthorpe wrote:
> I think this is another case of the HW can do it but the SW support is
> missing. IOMMU configuration and maybe firmware too, for instance.
Nope, not sure how you can make this leap. We've been specifically told
that peer-to-peer PCIe DMA is not
On 26/03/18 10:41 AM, Jason Gunthorpe wrote:
> On Mon, Mar 26, 2018 at 12:11:38PM +0100, Jonathan Cameron wrote:
>> On Tue, 13 Mar 2018 10:43:55 -0600
>> Logan Gunthorpe <log...@deltatee.com> wrote:
>>
>>> On 12/03/18 09:28 PM, Sinan Kaya wrote:
>>>
On 26/03/18 08:01 AM, Bjorn Helgaas wrote:
> On Mon, Mar 26, 2018 at 12:11:38PM +0100, Jonathan Cameron wrote:
>> On Tue, 13 Mar 2018 10:43:55 -0600
>> Logan Gunthorpe <log...@deltatee.com> wrote:
>>> It turns out that root ports that support P2P are far less commo
On 24/03/18 09:28 AM, Stephen Bates wrote:
> 1. There is no requirement for a single function to support internal DMAs but
> in the case of NVMe we do have a protocol specific way for a NVMe function to
> indicate it supports via the CMB BAR. Other protocols may also have such
> methods but
On 23/03/18 03:50 PM, Bjorn Helgaas wrote:
> Popping way up the stack, my original point was that I'm trying to
> remove restrictions on what devices can participate in peer-to-peer
> DMA. I think it's fairly clear that in conventional PCI, any devices
> in the same PCI hierarchy, i.e., below
On 21/03/18 03:27 AM, Christoph Hellwig wrote:
>> + const char *page, size_t count)
>> +{
>> +struct nvmet_port *port = to_nvmet_port(item);
>> +struct device *dev;
>> +struct pci_dev *p2p_dev = NULL;
>> +bool use_p2pmem;
>> +
>> +switch (page[0])
On 14/03/18 01:28 PM, Dan Williams wrote:
> P2P over PCI/PCI-X is quite common in devices like raid controllers.
> It would be useful if those configurations were not left behind so
> that Linux could feasibly deploy offload code to a controller in the
> PCI domain.
Thanks for the note. Neat.
On 14/03/18 12:51 PM, Bjorn Helgaas wrote:
> You are focused on PCIe systems, and in those systems, most topologies
> do have an upstream switch, which means two upstream bridges. I'm
> trying to remove that assumption because I don't think there's a
> requirement for it in the spec. Enforcing
On 14/03/18 06:16 AM, David Laight wrote:
> That surprises me (unless I missed something last time I read the spec).
> While P2P writes are relatively easy to handle, reads and any other TLP that
> require acks are a completely different proposition.
> There are no additional fields that can be
On 13/03/18 08:56 PM, Bjorn Helgaas wrote:
> I assume you want to exclude Root Ports because of multi-function
> devices and the "route to self" error. I was hoping for a reference
> to that so I could learn more about it.
I haven't been able to find where in the spec it forbids route to self.
On 13/03/18 05:19 PM, Sinan Kaya wrote:
> It is still a switch it can move packets but, maybe it can move data at
> 100kbps speed.
As Stephen pointed out, it's a requirement of the PCIe spec that a
switch supports P2P. If you want to sell a switch that does P2P with bad
performance then that's
On 13/03/18 05:08 PM, Bjorn Helgaas wrote:
> On Tue, Mar 13, 2018 at 10:31:55PM +, Stephen Bates wrote:
> If it *is* necessary because Root Ports and devices below them behave
> differently than in conventional PCI, I think you should include a
> reference to the relevant section of the
On 13/03/18 04:29 PM, Sinan Kaya wrote:
> If hardware doesn't support it, blacklisting should have been the right
> path and I still think that you should remove all switch business from the
> code.
> I did not hear enough justification for having a switch requirement
> for P2P.
I disagree.
>
On 13/03/18 03:22 PM, Sinan Kaya wrote:
> It sounds like you have very tight hardware expectations for this to work
> at this moment. You also don't want to generalize this code for others and
> address the shortcomings.
No, that's the way the community has pushed this work. Our original work
On 13/03/18 01:53 PM, Sinan Kaya wrote:
> I agree disabling globally would be bad. Somebody can always say I have
> ten switches on my system. I want to do peer-to-peer on one switch only. Now,
> this change weakened security for the other switches that I had no intention
> with doing P2P.
>
>
On 13/03/18 01:10 PM, Sinan Kaya wrote:
> I was thinking of this for the pci_p2pdma_add_client() case for the
> parent pointer.
>
> +struct pci_p2pdma_client {
> + struct list_head list;
> + struct pci_dev *client;
> + struct pci_dev *provider;
> +};
Yeah, that structure only
On 13/03/18 11:49 AM, Sinan Kaya wrote:
And there's also the ACS problem which means if you want to use P2P on the root
ports you'll have to disable ACS on the entire system. (Or preferably, the
IOMMU groups need to get more sophisticated to allow for dynamic changes).
Do you think you
On 12/03/18 09:28 PM, Sinan Kaya wrote:
Maybe, dev parameter should also be struct pci_dev so that you can get rid of
all to_pci_dev() calls in this code including find_parent_pci_dev() function.
No, this was mentioned in v2. find_parent_pci_dev is necessary because
the calling drivers
On 12/03/18 09:28 PM, Sinan Kaya wrote:
On 3/12/2018 3:35 PM, Logan Gunthorpe wrote:
Regarding the switch business, It is amazing how much trouble you went into
limit this functionality into very specific hardware.
I thought that we reached to an agreement that code would not impose
any
On 3/12/2018 1:41 PM, Jonathan Corbet wrote:
This all seems good, but...could we consider moving this documentation to
driver-api/PCI as it's converted to RST? That would keep it together with
similar materials and bring a bit more coherence to Documentation/ as a
whole.
Yup, I'll change this
-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/pci/Kconfig| 9 +
drivers/pci/p2pdma.c | 44
drivers/pci/pci.c | 6 ++
include/linux/pci-p2pdma.h | 5 +
4 files changed, 64 insertions(+)
diff
flag set.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Sagi Grimberg <s...@grimberg.me>
---
block/blk-core.c | 3 +++
include/linux/blk_types.h | 18 +-
include/linux/blkdev.h| 3 +++
3 files changed, 23 insertions(+), 1 deletion(-)
is P2P the entire SGL should be P2P.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/infiniband/core/rw.c | 13 +++--
1 file changed, 11 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/core/rw.c b/drivers/infiniband/core/rw.c
index c8963e
hch: partial rewrite of the initial code]
Signed-off-by: Christoph Hellwig <h...@lst.de>
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/nvme/target/configfs.c | 67 ++
drivers/nvme/target/core.c | 106 -
Register the CMB buffer as p2pmem and use the appropriate allocation
functions to create and destroy the IO SQ.
If the CMB supports WDS and RDS, publish it for use as P2P memory
by other devices.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/nvme/host/pci.
The DMA address used when mapping PCI P2P memory must be the PCI bus
address. Thus, introduce pci_p2pmem_[un]map_sg() to map the correct
addresses when using P2P memory.
For this, we assume that an SGL passed to these functions contain all
P2P memory or no P2P memory.
Signed-off-by: Logan
converted to restructured text
at this time.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Cc: Jonathan Corbet <cor...@lwn.net>
---
Documentation/PCI/index.rst | 14
Documentation/PCI/p2pdma.rst | 164 +++
Documentation/index.rst
new namespaces that are not supported
by that memory will fail.
Logan Gunthorpe (11):
PCI/P2PDMA: Support peer-to-peer memory
PCI/P2PDMA: Add sysfs group to display p2pmem stats
PCI/P2PDMA: Add PCI p2pmem dma mappings to adjust the bus offset
PCI/P2PDMA: Clear ACS P2P flags for all devices
Introduce a quirk to use CMB-like memory on older devices that have
an exposed BAR but do not advertise support for using CMBLOC and
CMBSIZE.
We'd like to use some of these older cards to test P2P memory.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Sagi Grimb
Add a sysfs group to display statistics about P2P memory that is
registered in each PCI device.
Attributes in the group display the total amount of P2P memory, the
amount available and whether it is published or not.
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
Documentati
-off-by: Logan Gunthorpe <log...@deltatee.com>
Reviewed-by: Sagi Grimberg <s...@grimberg.me>
---
drivers/nvme/host/core.c | 4
drivers/nvme/host/nvme.h | 1 +
drivers/nvme/host/pci.c | 19 +++
3 files changed, 20 insertions(+), 4 deletions(-)
diff --git a/drive
and feedback from Christoph
Hellwig.
Signed-off-by: Christoph Hellwig <h...@lst.de>
Signed-off-by: Logan Gunthorpe <log...@deltatee.com>
---
drivers/pci/Kconfig| 16 ++
drivers/pci/Makefile | 1 +
drivers/pci/p2pdma.c | 679 ++
On 05/03/18 05:49 PM, Oliver wrote:
It's in arch/powerpc/kernel/io.c as _memcpy_toio() and it has two full barriers!
Awesome!
Our io.h indicates that our iomem accessors are designed to provide x86ish
strong ordering of accesses to MMIO space. The git log indicates
arch/powerpc/kernel/io.c
On 05/03/18 03:28 PM, Bjorn Helgaas wrote:
If you put the #ifdef right here, then it's easier to read because we
can see that "oh, this is a special and uncommon case that I can
probably ignore".
Makes sense. I'll do that.
Thanks,
Logan
On 05/03/18 01:10 PM, Jason Gunthorpe wrote:
So when reading the above mlx code, we see the first wmb() being used
to ensure that CPU stores to cachable memory are visible to the DMA
triggered by the doorbell ring.
Oh, yes, that makes sense. Disregard my previous email as I was wrong.
Logan
On 05/03/18 12:57 PM, Sagi Grimberg wrote:
Keith, while we're on this, regardless of cmb, is SQE memcopy and DB
update ordering always guaranteed?
If you look at mlx4 (rdma device driver) that works exactly the same as
nvme you will find:
--
qp->sq.head += nreq;
On 05/03/18 11:02 AM, Sinan Kaya wrote:
writel has a barrier inside on ARM64.
https://elixir.bootlin.com/linux/latest/source/arch/arm64/include/asm/io.h#L143
Yes, and no barrier inside memcpy_toio as it uses __raw_writes. This
should be sufficient as we are only accessing addresses that
On 05/03/18 09:00 AM, Keith Busch wrote:
On Mon, Mar 05, 2018 at 12:33:29PM +1100, Oliver wrote:
On Thu, Mar 1, 2018 at 10:40 AM, Logan Gunthorpe <log...@deltatee.com> wrote:
@@ -429,10 +429,7 @@ static void __nvme_submit_cmd(struct nvme_queue *nvmeq,
{
u16 tail = nvmeq-&g
On 02/03/18 02:44 PM, Benjamin Herrenschmidt wrote:
Allright, so, I think I have a plan to fix this, but it will take a
little bit of time.
Basically the idea is to have firmware pass to Linux a region that's
known to not have anything in it that it can use for the vmalloc space
rather than
On 02/03/18 09:18 AM, Jason Gunthorpe wrote:
This allocator is already seems not useful for the P2P target memory
on a Mellanox NIC due to the way it has a special allocation flow
(windowing) and special usage requirements..
Nor can it be usefull for the doorbell memory in the NIC.
No one
On 01/03/18 05:36 PM, Dan Williams wrote:
On Thu, Mar 1, 2018 at 4:15 PM, Logan Gunthorpe <log...@deltatee.com> wrote:
On 01/03/18 10:44 AM, Bjorn Helgaas wrote:
I think these two statements are out of order, since the attributes
dereference pdev->p2pdma. And it looks like you s
On 01/03/18 10:44 AM, Bjorn Helgaas wrote:
I think these two statements are out of order, since the attributes
dereference pdev->p2pdma. And it looks like you set "error"
unnecessarily, since you return immediately looking at it.
Per the previous series, sysfs_create_group is must_check for
On 01/03/18 04:57 PM, Stephen Bates wrote:
We don't want to lump these all together without knowing which region you're
allocating from, right?
In all seriousness I do agree with you on these Keith in the long term. We
would consider adding property flags for the memory as it is added to
On 01/03/18 04:15 PM, Bjorn Helgaas wrote:
The question is what the relevant switch is. We call pci_enable_acs()
on every PCI device, including Root Ports. It looks like this relies
on get_upstream_bridge_port() to filter out some things. I don't
think get_upstream_bridge_port() is doing
On 01/03/18 04:26 PM, Benjamin Herrenschmidt wrote:
The big problem is not the vmemmap, it's the linear mapping.
Ah, yes, ok.
Logan
On 01/03/18 04:49 PM, Keith Busch wrote:
On Thu, Mar 01, 2018 at 11:00:51PM +, Stephen Bates wrote:
P2P is about offloading the memory and PCI subsystem of the host CPU
and this is achieved no matter which p2p_dev is used.
Even within a device, memory attributes for its various
On 01/03/18 04:20 PM, Jason Gunthorpe wrote:
On Thu, Mar 01, 2018 at 11:00:51PM +, Stephen Bates wrote:
No, locality matters. If you have a bunch of NICs and bunch of drives
and the allocator chooses to put all P2P memory on a single drive your
performance will suck horribly even if
On 01/03/18 04:00 PM, Benjamin Herrenschmidt wrote:
We use only 52 in practice but yes.
That's 64PB. If you use need
a sparse vmemmap for the entire space it will take 16TB which leaves you
with 63.98PB of address space left. (Similar calculations for other
numbers of address bits.)
We
1 - 100 of 179 matches
Mail list logo