| Age | Commit message (Collapse) | Author |
|---|
|
Pick up the first half of the RCH error handling series. The back half
needs some fixups for test regressions. Small conflicts with the PMU
work around register enumeration and setup helpers.
|
|
Pick up the recent fixes to how CPU caches are managed relative to
region setup / teardown, and make sure that all decoders transition
successfully before updating the region state from COMMIT => ACTIVE.
|
|
Switch-level (mid-level) decoders between the platform root and an
endpoint can dynamically switch modes between HDM-H and HDM-D[B]
depending on which region they target. Use the region type to fixup each
decoder that gets allocated to map the given region.
Note that endpoint decoders are meant to determine the region type, so
warn if those ever need to be fixed up, but since it is possible to
continue do so.
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168679262543.3436160.13053831955768440312.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
In preparation for support for HDM-D and HDM-DB configuration
(device-memory, and device-memory with back-invalidate). Rename the current
type designators to use HOSTONLYMEM and DEVMEM as a suffix.
HDM-DB can be supported by devices that are not accelerators, so DEVMEM is
a more generic term for that case.
Fixup one location where this type value was open coded.
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168679261369.3436160.7042443847605280593.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Jonathan reports that failed attempts to reset a region (teardown its
HDM decoder configuration) mistakenly advance the state of the region
to "not committed". Revert to the previous state of the region on reset
failure so that the reset can be re-attempted.
Reported-by: Jonathan Cameron <Jonathan.Cameron@Huawei.com>
Closes: http://lore.kernel.org/r/20230316171441.0000205b@Huawei.com
Fixes: 176baefb2eb5 ("cxl/hdm: Commit decoder state to hardware")
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168696507968.3590522.14484000711718573626.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
cxl_region_decode_reset() walks all the decoders associated with a given
region and disables them. Due to decoder ordering rules it is possible
that a switch in the topology notices that a given decoder can not be
shutdown before another region with a higher HPA is shutdown first. That
can leave the region in a partially committed state.
Capture that state in a new CXL_REGION_F_NEEDS_RESET flag and require
that a successful cxl_region_decode_reset() attempt must be completed
before cxl_region_probe() accepts the region.
This is a corollary for the bug that Jonathan identified in "CXL/region
: commit reset of out of order region appears to succeed." [1].
Cc: Jonathan Cameron <Jonathan.Cameron@Huawei.com>
Link: http://lore.kernel.org/r/20230316171441.0000205b@Huawei.com [1]
Fixes: 176baefb2eb5 ("cxl/hdm: Commit decoder state to hardware")
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168696507423.3590522.16254212607926684429.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Vikram raised a concern with the theoretical case of a CPU sending
MemClnEvict to a device that is not prepared to receive. MemClnEvict is
a message that is sent after a CPU has taken ownership of a cacheline
from accelerator memory (HDM-DB). In the case of hotplug or HDM decoder
reconfiguration it is possible that the CPU is holding old contents for
a new device that has taken over the physical address range being cached
by the CPU.
To avoid this scenario, invalidate caches prior to tearing down an HDM
decoder configuration.
Now, this poses another problem that it is possible for something to
speculate into that space while the decode configuration is still up, so
to close that gap also invalidate prior to establish new contents behind
a given physical address range.
With this change the cache invalidation is now explicit and need not be
checked in cxl_region_probe(), and that obviates the need for
CXL_REGION_F_INCOHERENT.
Cc: Jonathan Cameron <Jonathan.Cameron@Huawei.com>
Fixes: d18bc74aced6 ("cxl/region: Manage CPU caches relative to DPA invalidation events")
Reported-by: Vikram Sethi <vsethi@nvidia.com>
Closes: http://lore.kernel.org/r/BYAPR12MB33364B5EB908BF7239BB996BBD53A@BYAPR12MB3336.namprd12.prod.outlook.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168696506886.3590522.4597053660991916591.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
For symmetry with the recent rename of ->dport_dev for a 'struct
cxl_dport', add the "_dev" suffix to the ->uport property of a 'struct
cxl_port'. These devices represent the downstream-port-device and
upstream-port-device respectively in the CXL/PCIe topology.
Signed-off-by: Terry Bowman <terry.bowman@amd.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/20230622205523.85375-6-terry.bowman@amd.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Reading code like dport->dport does not immediately suggest that this
points to the corresponding device structure of the dport. Rename
struct member @dport to @dport_dev.
While at it, also rename @new argument of add_dport() to @dport. This
better describes the variable as a dport (e.g. new->dport becomes to
dport->dport_dev).
Co-developed-by: Terry Bowman <terry.bowman@amd.com>
Signed-off-by: Terry Bowman <terry.bowman@amd.com>
Signed-off-by: Robert Richter <rrichter@amd.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/20230622205523.85375-5-terry.bowman@amd.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
User space may need to know which region, if any, maps the poison
address(es) logged in a cxl_poison trace event. Since the mapping
of DPAs (device physical addresses) to a region can change, the
kernel must provide this information at the time the poison list
is read. The event informs user space that at event <timestamp>
this <region> mapped to this <DPA>, which is poisoned.
The cxl_poison trace event is already wired up to log the region
name and uuid if it receives param 'struct cxl_region'.
In order to provide that cxl_region, add another method for gathering
poison - by committed endpoint decoder mappings. This method is only
available with CONFIG_CXL_REGION and is only used if a region actually
maps the memdev where poison is being read. After the region driver
reads the poison list for all the mapped resources, poison is read for
any remaining unmapped resources.
The default method remains: read the poison by memdev resource.
Signed-off-by: Alison Schofield <alison.schofield@intel.com>
Tested-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/438b01ccaa70592539e8eda4eb2b1d617ba03160.1681838292.git.alison.schofield@intel.com
Tested-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Each time the contents of a given HPA are potentially changed in a cache
incoherent manner the CXL core sets CXL_REGION_F_INCOHERENT to
invalidate CPU caches before the region is used.
Successful invocation of attach_target() indicates that DPA has been
newly assigned to a given HPA in the dynamic region creation flow.
However, attach_target() is also reused in the autodiscovery flow where
the region was activated by platform firmware. In that case there is no
need to invalidate caches because that region is already in active use
and nothing about the autodiscovery flow modifies the HPA-to-DPA
relationship.
In the autodiscovery case cxl_region_attach() exits early after
determining the endpoint decoder is already correctly attached to the
region.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Reviewed-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168002858817.50647.1217607907088920888.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
RCDs (CXL memory devices that link train without VH capability and show
up as root complex integrated endpoints), hide the presence of the link
between the endpoint and the host-bridge. The CXL region setup/teardown
paths assume that a link hop is present and go looking for at least one
'struct cxl_port' instance between the CXL root port-object and an
endpoint port-object leading to crashes of the form:
BUG: kernel NULL pointer dereference, address: 0000000000000008
[..]
RIP: 0010:cxl_region_setup_targets+0x3e9/0xae0 [cxl_core]
[..]
Call Trace:
<TASK>
cxl_region_attach+0x46c/0x7a0 [cxl_core]
cxl_create_region+0x20b/0x270 [cxl_core]
cxl_mock_mem_probe+0x641/0x800 [cxl_mock_mem]
platform_probe+0x5b/0xb0
Detect RCDs explicitly and skip walking the non-existent port hierarchy
between root and endpoint in that case.
While this has been a problem since:
commit 0a19bfc8de93 ("cxl/port: Add RCD endpoint port enumeration")
...it becomes a more reliable crash scenario with the new autodiscovery
implementation.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168002858268.50647.728091521032131326.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
The find_cxl_root() helper is used to lookup root decoders and other CXL
platform topology information for a given endpoint. It turns out that
for RCDs it has never worked. The result of find_cxl_root(&cxlmd->dev)
is always NULL for the RCH topology case because it expects to find a
cxl_port at the host-bridge. RCH topologies only have the root cxl_port
object with the host-bridge as a dport. While there are no reports of
this being a problem to date, by inspection region enumeration should
crash as a result of this problem, and it does in a local unit test for
this scenario.
However, an observation that ever since:
commit f17b558d6663 ("cxl/pmem: Refactor nvdimm device registration, delete the workqueue")
...all callers of find_cxl_root() occur after the memdev connection to
the port topology has been established. That means that find_cxl_root()
can be simplified to a walk of the endpoint port topology to the root.
Switch to that arrangement which also fixes the RCD bug.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/168002857715.50647.344876437247313909.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Pick up some fixes from exposure of for-6.3/cxl-ram-region in
linux-next.
|
|
The new cxl_add_to_region() function returns an uninitialized
value on success:
drivers/cxl/core/region.c:2628:6: error: variable 'rc' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized]
if (IS_ERR(cxlr)) {
^~~~~~~~~~~~
drivers/cxl/core/region.c:2654:9: note: uninitialized use occurs here
return rc;
Simplify the logic to have the rc variable always initialized in the
same place.
Fixes: a32320b71f08 ("cxl/region: Add region autodiscovery")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20230213101220.3821689-1-arnd@kernel.org
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Include the support for enumerating and provisioning ram regions for
v6.3. This also include a default policy change for ram / volatile
device-dax instances to assign them to the dax_kmem driver by default.
|
|
While platform firmware takes some responsibility for mapping the RAM
capacity of CXL devices present at boot, the OS is responsible for
mapping the remainder and hot-added devices. Platform firmware is also
responsible for identifying the platform general purpose memory pool,
typically DDR attached DRAM, and arranging for the remainder to be 'Soft
Reserved'. That reservation allows the CXL subsystem to route the memory
to core-mm via memory-hotplug (dax_kmem), or leave it for dedicated
access (device-dax).
The new 'struct cxl_dax_region' object allows for a CXL memory resource
(region) to be published, but also allow for udev and module policy to
act on that event. It also prevents cxl_core.ko from having a module
loading dependency on any drivers/dax/ modules.
Tested-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/167602003896.1924368.10335442077318970468.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Region autodiscovery is an asynchronous state machine advanced by
cxl_port_probe(). After the decoders on an endpoint port are enumerated
they are scanned for actively enabled instances. Each active decoder is
flagged for auto-assembly CXL_DECODER_F_AUTO and attached to a region.
If a region does not already exist for the address range setting of the
decoder one is created. That creation process may race with other
decoders of the same region being discovered since cxl_port_probe() is
asynchronous. A new 'struct cxl_root_decoder' lock, @range_lock, is
introduced to mitigate that race.
Once all decoders have arrived, "p->nr_targets == p->interleave_ways",
they are sorted by their relative decode position. The sort algorithm
involves finding the point in the cxl_port topology where one leg of the
decode leads to deviceA and the other deviceB. At that point in the
topology the target order in the 'struct cxl_switch_decoder' indicates
the relative position of those endpoint decoders in the region.
>From that point the region goes through the same setup and validation
steps as user-created regions, but instead of programming the decoders
it validates that driver would have written the same values to the
decoders as were already present.
Tested-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/167601999958.1924368.9366954455835735048.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
In preparation for region autodiscovery, that needs all devices
discovered before their relative position in the region can be
determined, consolidate all position dependent validation in a helper.
Recall that in the on-demand region creation flow the end-user picks the
position of a given endpoint decoder in a region. In the autodiscovery
case the position of an endpoint decoder can only be determined after
all other endpoint decoders that claim to decode the region's address
range have been enumerated and attached. So, in the autodiscovery case
endpoint decoders may be attached before their relative position is
known. Once all decoders arrive, then positions can be determined and
validated with cxl_region_validate_position() the same as user initiated
on-demand creation.
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Tested-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/167601997584.1924368.4615769326126138969.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Jonathan noticed that the target list setup is not unwound completely
upon error. Undo all the setup in the 'err_decrement:' exit path.
Fixes: 27b3f8d13830 ("cxl/region: Program target lists")
Reported-by: Jonathan Cameron <Jonathan.Cameron@Huawei.com>
Link: http://lore.kernel.org/r/20230208123031.00006990@Huawei.com
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/167601996980.1924368.390423634911157277.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Region autodiscovery is the process of kernel creating 'struct
cxl_region' object to represent active CXL memory ranges it finds
already active in hardware when the driver loads. Typically this happens
when platform firmware establishes CXL memory regions and then publishes
them in the memory map. However, this can also happen in the case of
kexec-reboot after the kernel has created regions.
In the autodiscovery case the region creation process starts with a
known endpoint decoder. Refactor attach_target() into a helper that is
suitable to be called from either sysfs, for runtime region creation, or
from cxl_port_probe() after it has enumerated all endpoint decoders.
The cxl_port_probe() context is an async device-core probing context, so
it is not appropriate to allow SIGTERM to interrupt the assembly
process. Refactor attach_target() to take @cxled and @state as arguments
where @state indicates whether waiting from the region rwsem is
interruptible or not.
No behavior change is intended.
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Fan Ni <fan.ni@samsung.com>
Link: https://lore.kernel.org/r/167601996393.1924368.2202255054618600069.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Expand the region creation infrastructure to enable 'ram'
(volatile-memory) regions. The internals of create_pmem_region_store()
and create_pmem_region_show() are factored out into helpers
__create_region() and __create_region_show() for the 'ram' case to
reuse.
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Gregory Price <gregory.price@memverge.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Fan Ni <fan.ni@samsung.com>
Link: https://lore.kernel.org/r/167601995775.1924368.352616146815830591.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
In preparation for a new region mode, do not, for example, allow
'ram' decoders to be assigned to 'pmem' regions and vice versa.
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Gregory Price <gregory.price@memverge.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Fan Ni <fan.ni@samsung.com>
Link: https://lore.kernel.org/r/167601995111.1924368.7459128614177994602.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Shipping versions of the cxl-cli utility expect all regions to have a
'uuid' attribute. In preparation for 'ram' regions, update the 'uuid'
attribute to return an empty string which satisfies the current
expectations of 'cxl list -R'. Otherwise, 'cxl list -R' fails in the
presence of regions with the 'uuid' attribute missing. Force the
attribute to be read-only as there is no facility or expectation for a
'ram' region to recall its uuid from one boot to the next.
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Tested-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/167601994558.1924368.12612811533724694444.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
In preparation for a new region type, "ram" regions, add a mode
attribute to clarify the mode of the decoders that can be added to a
region. Share the internals of mode_show() (for decoders) with the
region case.
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Gregory Price <gregory.price@memverge.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Fan Ni <fan.ni@samsung.com>
Link: https://lore.kernel.org/r/167601993930.1924368.4305018565539515665.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Merge the general CXL updates with fixes targeting v6.2-rc for v6.3.
Resolve a conflict with the fix and move of cxl_report_and_clear() from
pci.c to core/pci.c.
|
|
A passthrough decoder is a decoder that maps only 1 target. It is a
special case because it does not impose any constraints on the
interleave-math as compared to a decoder with multiple targets. Extend
the passthrough case to multi-target-capable decoders that only have one
target selected. I.e. the current code was only considering passthrough
*ports* which are only a subset of the potential passthrough decoder
scenarios.
Fixes: e4f6dfa9ef75 ("cxl/region: Fix 'distance' calculation with passthrough ports")
Cc: <stable@vger.kernel.org>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/167564540422.847146.13816934143225777888.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Not all decoders have a reset callback.
The CXL specification allows a host bridge with a single root port to
have no explicit HDM decoders. Currently the region driver assumes there
are none. As such the CXL core creates a special pass through decoder
instance without a commit/reset callback.
Prior to this patch, the ->reset() callback was called unconditionally when
calling cxl_region_decode_reset. Thus a configuration with 1 Host Bridge,
1 Root Port, and one directly attached CXL type 3 device or multiple CXL
type 3 devices attached to downstream ports of a switch can cause a null
pointer dereference.
Before the fix, a kernel crash was observed when we destroy the region, and
a pass through decoder is reset.
The issue can be reproduced as below,
1) create a region with a CXL setup which includes a HB with a
single root port under which a memdev is attached directly.
2) destroy the region with cxl destroy-region regionX -f.
Fixes: 176baefb2eb5 ("cxl/hdm: Commit decoder state to hardware")
Cc: <stable@vger.kernel.org>
Signed-off-by: Fan Ni <fan.ni@samsung.com>
Reviewed-by: Davidlohr Bueso <dave@stgolabs.net>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Tested-by: Gregory Price <gregory.price@memverge.com>
Reviewed-by: Gregory Price <gregory.price@memverge.com>
Link: https://lore.kernel.org/r/20221215170909.2650271-1-fan.ni@samsung.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Correct spelling mistakes (reported by codespell).
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Cc: Alison Schofield <alison.schofield@intel.com>
Cc: Vishal Verma <vishal.l.verma@intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Ben Widawsky <bwidawsk@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: linux-cxl@vger.kernel.org
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Alison Schofield <alison.schofield@intel.com>
Link: https://lore.kernel.org/r/20230125032221.21277-1-rdunlap@infradead.org
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Both cxl_switch_decoders() and cxl_endpoint_decoders() are considered by
cxl_region_decode_commit(). Flag cases where cxl_switch_decoders with
multiple targets, or cxl_endpoint_decoders do not have a commit callback
set. The switch case is unlikely to happen since switches are only
enumerated by the CXL core, but the endpoint case may support decoders
defined by drivers outside of drivers/cxl, like accerator drivers.
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/167124081824.1626103.1555704405392757219.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
No need for more than once per module load.
Signed-off-by: Davidlohr Bueso <dave@stgolabs.net>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/20221215183836.24136-1-dave@stgolabs.net
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Due to a typo, the check of whether or not a memdev has already been
used as a target for the region (above code piece) will always be
skipped. Given a memdev with more than one HDM decoder, an interleaved
region can be created that maps multiple HPAs to the same DPA. According
to CXL spec 3.0 8.1.3.8.4, "Aliasing (mapping more than one Host
Physical Address (HPA) to a single Device Physical Address) is
forbidden."
Fix this by using existing iterator for memdev reuse check.
Cc: <stable@vger.kernel.org>
Fixes: 384e624bb211 ("cxl/region: Attach endpoint decoders")
Signed-off-by: Fan Ni <fan.ni@samsung.com>
Link: https://lore.kernel.org/r/20221107212153.745993-1-fan.ni@samsung.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Change names for interleave ways macros to clearly indicate which
variable is encoded and which is the actual ways value.
ways == interleave ways
eiw == encoded interleave ways
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/167027516228.3124679.11265039496968588580.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Change names for granularity macros to clearly indicate which
variable is encoded and which is the actual granularity.
granularity == interleave granularity
eig == encoded interleave granularity
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Alison Schofield <alison.schofield@intel.com>
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/167027493237.3124429.8948852388671827664.stgit@djiang5-desk3.ch.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
There is a spelling mistake in a dev_warn message. Fix it.
Signed-off-by: Colin Ian King <colin.i.king@gmail.com>
Link: https://lore.kernel.org/r/20221205091819.1943564-1-colin.i.king@gmail.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Pick CXL PMEM security commands for v6.2. Resolve conflicts with the
removal of the cxl_pmem_wq.
|
|
A "DPA invalidation event" is any scenario where the contents of a DPA
(Device Physical Address) is modified in a way that is incoherent with
CPU caches, or if the HPA (Host Physical Address) to DPA association
changes due to a remapping event.
PMEM security events like Unlock and Passphrase Secure Erase already
manage caches through LIBNVDIMM, so that leaves HPA to DPA remap events
that need cache management by the CXL core. Those only happen when the
boot time CXL configuration has changed. That event occurs when
userspace attaches an endpoint decoder to a region configuration, and
that region is subsequently activated.
The implications of not invalidating caches between remap events is that
reads from the region at different points in time may return different
results due to stale cached data from the previous HPA to DPA mapping.
Without a guarantee that the region contents after cxl_region_probe()
are written before being read (a layering-violation assumption that
cxl_region_probe() can not make) the CXL subsystem needs to ensure that
reads that precede writes see consistent results.
A CONFIG_CXL_REGION_INVALIDATION_TEST option is added to support debug
and unit testing of the CXL implementation in QEMU or other environments
where cpu_cache_has_invalidate_memregion() returns false. This may prove
too restrictive for QEMU where the HDM decoders are emulated, but in
that case the CXL subsystem needs some new mechanism / indication that
the HDM decoder is emulated and not a passthrough of real hardware.
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/166993222098.1995348.16604163596374520890.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
cxl_region_probe() allows for regions not in the 'commit' state to be
enabled. Fail probe when the region is not committed otherwise the
kernel may indicate that an address range is active when none of the
decoders are active.
Fixes: 8d48817df6ac ("cxl/region: Add region driver boiler plate")
Cc: <stable@vger.kernel.org>
Reviewed-by: Davidlohr Bueso <dave@stgolabs.net>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/166993220462.1995348.1698008475198427361.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
The three objects 'struct cxl_nvdimm_bridge', 'struct cxl_nvdimm', and
'struct cxl_pmem_region' manage CXL persistent memory resources. The
bridge represents base platform resources, the nvdimm represents one or
more endpoints, and the region is a collection of nvdimms that
contribute to an assembled address range.
Their relationship is such that a region is torn down if any component
endpoints are removed. All regions and endpoints are torn down if the
foundational bridge device goes down.
A workqueue was deployed to manage these interdependencies, but it is
difficult to reason about, and fragile. A recent attempt to take the CXL
root device lock in the cxl_mem driver was reported by lockdep as
colliding with the flush_work() in the cxl_pmem flows.
Instead of the workqueue, arrange for all pmem/nvdimm devices to be torn
down immediately and hierarchically. A similar change is made to both
the 'cxl_nvdimm' and 'cxl_pmem_region' objects. For bisect-ability both
changes are made in the same patch which unfortunately makes the patch
bigger than desired.
Arrange for cxl_memdev and cxl_region to register a cxl_nvdimm and
cxl_pmem_region as a devres release action of the bridge device.
Additionally, include a devres release action of the cxl_memdev or
cxl_region device that triggers the bridge's release action if an endpoint
exits before the bridge. I.e. this allows either unplugging the bridge,
or unplugging and endpoint to result in the same cleanup actions.
To keep the patch smaller the cleanup of the now defunct workqueue
infrastructure is saved for a follow-on patch.
Tested-by: Robert Richter <rrichter@amd.com>
Link: https://lore.kernel.org/r/166993041773.1882361.16444301376147207609.stgit@dwillia2-xfh.jf.intel.com
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
At region creation time the next region-id is atomically cached so that
there is predictability of region device names. If that region is
destroyed and then a new one is created the region id increments. That
ends up looking like a memory leak, or is otherwise surprising that
identifiers roll forward even after destroying all previously created
regions.
Try to reuse rather than free old region ids at region release time.
While this fixes a cosmetic issue, the needlessly advancing memory
region-id gives the appearance of a memory leak, hence the "Fixes" tag,
but no "Cc: stable" tag.
Cc: Ben Widawsky <bwidawsk@kernel.org>
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Fixes: 779dd20cfb56 ("cxl/region: Add region creation support")
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Link: https://lore.kernel.org/r/166752186062.947915.13200195701224993317.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
When programming port decode targets, the algorithm wants to ensure that
two devices are compatible to be programmed as peers beneath a given
port. A compatible peer is a target that shares the same dport, and
where that target's interleave position also routes it to the same
dport. Compatibility is determined by the device's interleave position
being >= to distance. For example, if a given dport can only map every
Nth position then positions less than N away from the last target
programmed are incompatible.
The @distance for the host-bridge's cxl_port in a simple dual-ported
host-bridge configuration with 2 direct-attached devices is 1, i.e. An
x2 region divided by 2 dports to reach 2 region targets.
An x4 region under an x2 host-bridge would need 2 intervening switches
where the @distance at the host bridge level is 2 (x4 region divided by
2 switches to reach 4 devices).
However, the distance between peers underneath a single ported
host-bridge is always zero because there is no limit to the number of
devices that can be mapped. In other words, there are no decoders to
program in a passthrough, all descendants are mapped and distance only
starts matters for the intervening descendant ports of the passthrough
port.
Add tracking for the number of dports mapped to a port, and use that to
detect the passthrough case for calculating @distance.
Cc: <stable@vger.kernel.org>
Reported-by: Bobo WL <lmw.bobo@gmail.com>
Reported-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: http://lore.kernel.org/r/20221010172057.00001559@huawei.com
Fixes: 27b3f8d13830 ("cxl/region: Program target lists")
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Link: https://lore.kernel.org/r/166752185440.947915.6617495912508299445.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
When a region is deleted any targets that have been previously assigned
to that region hold references to it. Trigger those references to
drop by detaching all targets at unregister_region() time.
Otherwise that region object will leak as userspace has lost the ability
to detach targets once region sysfs is torn down.
Cc: <stable@vger.kernel.org>
Fixes: b9686e8c8e39 ("cxl/region: Enable the assignment of endpoint decoders to regions")
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Link: https://lore.kernel.org/r/166752183055.947915.17681995648556534844.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Some regions may not have any address space allocated. Skip them when
validating HPA order otherwise a crash like the following may result:
devm_cxl_add_region: cxl_acpi cxl_acpi.0: decoder3.4: created region9
BUG: kernel NULL pointer dereference, address: 0000000000000000
[..]
RIP: 0010:store_targetN+0x655/0x1740 [cxl_core]
[..]
Call Trace:
<TASK>
kernfs_fop_write_iter+0x144/0x200
vfs_write+0x24a/0x4d0
ksys_write+0x69/0xf0
do_syscall_64+0x3a/0x90
store_targetN+0x655/0x1740:
alloc_region_ref at drivers/cxl/core/region.c:676
(inlined by) cxl_port_attach_region at drivers/cxl/core/region.c:850
(inlined by) cxl_region_attach at drivers/cxl/core/region.c:1290
(inlined by) attach_target at drivers/cxl/core/region.c:1410
(inlined by) store_targetN at drivers/cxl/core/region.c:1453
Cc: <stable@vger.kernel.org>
Fixes: 384e624bb211 ("cxl/region: Attach endpoint decoders")
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/166752182461.947915.497032805239915067.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
When an intermediate port's decoders have been exhausted by existing
regions, and creating a new region with the port in question in it's
hierarchical path is attempted, cxl_port_attach_region() fails to find a
port decoder (as would be expected), and drops into the failure / cleanup
path.
However, during cleanup of the region reference, a sanity check attempts
to dereference the decoder, which in the above case didn't exist. This
causes a NULL pointer dereference BUG.
To fix this, refactor the decoder allocation and de-allocation into
helper routines, and in this 'free' routine, check that the decoder,
@cxld, is valid before attempting any operations on it.
Cc: <stable@vger.kernel.org>
Suggested-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Fixes: 384e624bb211 ("cxl/region: Attach endpoint decoders")
Link: https://lore.kernel.org/r/20221101074100.1732003-1-vishal.l.verma@intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Not all decoders have a commit callback.
The CXL specification allows a host bridge with a single root port to
have no explicit HDM decoders. Currently the region driver assumes there
are none. As such the CXL core creates a special pass through decoder
instance without a commit callback.
Prior to this patch, the ->commit() callback was called unconditionally.
Thus a configuration with 1 Host Bridge, 1 Root Port, 1 switch with
multiple downstream ports below which there are multiple CXL type 3
devices results in a situation where committing the region causes a null
pointer dereference.
Reported-by: Bobo WL <lmw.bobo@gmail.com>
Fixes: 176baefb2eb5 ("cxl/hdm: Commit decoder state to hardware")
Signed-off-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Link: https://lore.kernel.org/r/20220818164210.2084-1-Jonathan.Cameron@huawei.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
The endpoint decode granularity must be <= the window granularity
otherwise capacity in the endpoints is lost in the decode. Consider an
attempt to have a region granularity of 512 with 4 devices within a
window that maps 2 host bridges at a granularity of 256 bytes:
HPA DPA Offset HB Port EP
0x0 0x0 0 0 0
0x100 0x0 1 0 2
0x200 0x100 0 0 0
0x300 0x100 1 0 2
0x400 0x200 0 1 1
0x500 0x200 1 1 3
0x600 0x300 0 1 1
0x700 0x300 1 1 3
0x800 0x400 0 0 0
0x900 0x400 1 0 2
0xA00 0x500 0 0 0
0xB00 0x500 1 0 2
Notice how endpoint0 maps HPA 0x0 and 0x200 correctly, but then at HPA
0x800 it results in DPA 0x200-0x400 on being skipped.
Fix this by restricing the region granularity to be equal to the window
granularity resulting in the following for a x4 region under a x2 window
at a granularity of 256.
HPA DPA Offset HB Port EP
0x0 0x0 0 0 0
0x100 0x0 1 0 2
0x200 0x0 0 1 1
0x300 0x0 1 1 3
0x400 0x100 0 0 0
0x500 0x100 1 0 2
0x600 0x100 0 1 1
0x700 0x100 1 1 3
Not that it ever made practical sense to support region granularity >
window granularity. The window rotates host bridges causing endpoints to
never see a consecutive stream of requests at the desired granularity
without breaks to issue cycles to the other host bridge.
Fixes: 80d10a6cee05 ("cxl/region: Add interleave geometry attributes")
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Link: https://lore.kernel.org/r/165973127171.1526540.9923273539049172976.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
In cases where the decode fans out as it traverses downstream, the
interleave granularity needs to increment to identify the port selector
bits out of the remaining address bits. For example, recall that with an
x2 parent port intereleave (IW == 1), the downstream decode for children
of those ports will either see address bit IG+8 always set, or address
bit IG+8 always clear. So if the child port needs to select a downstream
port it can only use address bits starting at IG+9 (where IG and IW are
the CXL encoded values for interleave granularity (ilog2(ig) - 8) and
ways (ilog2(iw))).
When the parent port interleave is x1 no such masking occurs and the
child port can maintain the granularity that was routed to the parent
port.
Reported-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Link: https://lore.kernel.org/r/165973126583.1526540.657948655360009242.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
A recent bug fix added the setup of the endpoint decoder interleave
geometry settings to cxl_region_attach(). Move the HPA setup there as
well to keep all endpoint decoder parameter setting in a central
location.
For symmetry, move endpoint HPA teardown to cxl_region_detach(), and for
switches move HPA setup / teardown to cxl_port_{setup,reset}_targets().
Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Vishal Verma <vishal.l.verma@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Link: https://lore.kernel.org/r/165973126020.1526540.14701949254436069807.stgit@dwillia2-xfh.jf.intel.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Jonathan notes:
"Curiously interleave ways = 1 for the EPs which is obviously wrong"
...while testing the latest CXL development branch on QEMU.
It turns out the region creation process failed to program the endpoint
decoders. This was missed because the default settings of x1 at 4K
intereleave still results in the region appearing to function. Jonathan
caught the bug by reverse mapping the translations that need to happen
for the QEMU support.
Link: https://lore.kernel.org/r/62e95fdf9f6e2_30440294e4@dwillia2-xfh.jf.intel.com.notmuch
Fixes: 384e624bb211 ("cxl/region: Attach endpoint decoders")
Reported-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/165951146336.967013.11160153960900111443.stgit@dwillia2-xfh.jf.intel.com
Acked-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
|
Sphinx reported indentation warnings:
Documentation/driver-api/cxl/memory-devices:457: ./drivers/cxl/core/region.c:732: WARNING: Unexpected indentation.
Documentation/driver-api/cxl/memory-devices:457: ./drivers/cxl/core/region.c:733: WARNING: Block quote ends without a blank line; unexpected unindent.
Documentation/driver-api/cxl/memory-devices:457: ./drivers/cxl/core/region.c:735: WARNING: Unexpected indentation.
These warnings above are due to missing blank line padding in the nested list
in kernel-doc comment for cxl_rr_ep_add().
Add the paddings to fix the warnings.
Fixes: 384e624bb211b4 ("cxl/region: Attach endpoint decoders")
Signed-off-by: Bagas Sanjaya <bagasdotme@gmail.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Link: https://lore.kernel.org/r/20220804075448.98241-2-bagasdotme@gmail.com
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
|
