- 11 Feb, 2023 26 commits
-
-
Dan Williams authored
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.
-
Dan Williams authored
Pick up some final miscellaneous updates for v6.3 including support for communicating 'exclusive' and 'enabled' state of commands.
-
Ira Weiny authored
The command comment had grammatical errors. In an attempt to fix those it was noted that the comment and the query command were not in sync. Now that the query command returns excluded and device unsupported command information. Update the kdoc and fix the grammatical errors. [1] https://lore.kernel.org/all/63b4ec4e37cc1_5178e2941d@dwillia2-xfh.jf.intel.com.notmuch/Reviewed-by:
Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by:
Ira Weiny <ira.weiny@intel.com> Link: https://lore.kernel.org/r/20221222-cxl-misc-v4-4-62f701c1cdd1@intel.comSigned-off-by:
Dan Williams <dan.j.williams@intel.com>
-
Ira Weiny authored
It was pointed out that commands not supported by the device or excluded by the kernel were being returned in cxl_query_cmd().[1] While libcxl correctly handles failing commands, it is more efficient to not issue an invalid command in the first place. This can't be done without additional information being returned from cxl_query_cmd(). In addition, information about the availability of commands can be useful for debugging. Add flags to struct cxl_command_info which reflect if a command is enabled and/or exclusive to the kernel. [1] https://lore.kernel.org/all/63b4ec4e37cc1_5178e2941d@dwillia2-xfh.jf.intel.com.notmuch/Suggested-by:
-
Ira Weiny authored
The CXL command enum is exported to user space and must maintain backwards compatibility. Add comment that new defines must be added to the end of the list. Suggested-by:
Dave Jiang <dave.jiang@intel.com> Signed-off-by:
-
Ira Weiny authored
CXL_CMD_FLAG_NONE is not used, remove it. Reviewed-by:
-
Dan Williams authored
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:
-
Dan Williams authored
The default mode for device-dax instances is backwards for RAM-regions as evidenced by the fact that it tends to catch end users by surprise. "Where is my memory?". Recall that platforms are increasingly shipping with performance-differentiated memory pools beyond typical DRAM and NUMA effects. This includes HBM (high-bandwidth-memory) and CXL (dynamic interleave, varied media types, and future fabric attached possibilities). For this reason the EFI_MEMORY_SP (EFI Special Purpose Memory => Linux 'Soft Reserved') attribute is expected to be applied to all memory-pools that are not the general purpose pool. This designation gives an Operating System a chance to defer usage of a memory pool until later in the boot process where its performance properties can be interrogated and administrator policy can be applied. 'Soft Reserved' memory can be anything from too limited and precious to be part of the general purpose pool (HBM), too slow to host hot kernel data structures (some PMEM media), or anything in between. However, in the absence of an explicit policy, the memory should at least be made usable by default. The current device-dax default hides all non-general-purpose memory behind a device interface. The expectation is that the distribution of users that want the memory online by default vs device-dedicated-access by default follows the Pareto principle. A small number of enlightened users may want to do userspace memory management through a device, but general users just want the kernel to make the memory available with an option to get more advanced later. Arrange for all device-dax instances not backed by PMEM to default to attaching to the dax_kmem driver. From there the baseline memory hotplug policy (CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE / memhp_default_state=) gates whether the memory comes online or stays offline. Where, if it stays offline, it can be reliably converted back to device-mode where it can be partitioned, or fronted by a userspace allocator. So, if someone wants device-dax instances for their 'Soft Reserved' memory: 1/ Build a kernel with CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE=n or boot with memhp_default_state=offline, or roll the dice and hope that the kernel has not pinned a page in that memory before step 2. 2/ Write a udev rule to convert the target dax device(s) from 'system-ram' mode to 'devdax' mode: daxctl reconfigure-device $dax -m devdax -f Cc: Michal Hocko <mhocko@suse.com> Cc: David Hildenbrand <david@redhat.com> Cc: Dave Hansen <dave.hansen@linux.intel.com> Reviewed-by:
Gregory Price <gregory.price@memverge.com> Tested-by:
-
Dan Williams authored
In preparation for the CXL region driver to take over the responsibility of registering device-dax instances for CXL regions, move the registration of "hmem" devices to dax_hmem.ko. Previously the builtin component of this enabling (drivers/dax/hmem/device.o) would register platform devices for each address range and trigger the dax_hmem.ko module to load and attach device-dax instances to those devices. Now, the ranges are collected from the HMAT and EFI memory map walking, but the device creation is deferred. A new "hmem_platform" device is created which triggers dax_hmem.ko to load and register the platform devices. Tested-by:
-
Dan Williams authored
In preparation for hmem platform devices to be unregistered, stop using platform_device_add_resources() to convey the address range. The platform_device_add_resources() API causes an existing "Soft Reserved" iomem resource to be re-parented under an inserted platform device resource. When that platform device is deleted it removes the platform device resource and all children. Instead, it is sufficient to convey just the address range and let request_mem_region() insert resources to indicate the devices active in the range. This allows the "Soft Reserved" resource to be re-enumerated upon the next probe event. Reviewed-by:
-
Dan Williams authored
Empty driver remove callbacks can just be elided. Reviewed-by:
Vishal Verma <vishal.l.verma@intel.com> Reviewed-by:
-
Dan Williams authored
In preparation for moving more filtering of "hmem" ranges into the dax_hmem.ko module, update the initcall levels. HMAT range registration moves to subsys_initcall() to be done before Soft Reservation probing, and Soft Reservation probing is moved to device_initcall() to be done before dax_hmem.ko initialization if it is built-in. Tested-by:
-
Dan Williams authored
Take two endpoints attached to the first switch on the first host-bridge in the cxl_test topology and define a pre-initialized region. This is a x2 interleave underneath a x1 CXL Window. $ modprobe cxl_test $ # cxl list -Ru { "region":"region3", "resource":"0xf010000000", "size":"512.00 MiB (536.87 MB)", "interleave_ways":2, "interleave_granularity":4096, "decode_state":"commit" } Tested-by: -
Dan Williams authored
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:
-
Dan Williams authored
Jonathan points out that the shared code between the switch and endpoint case is small. Before adding another is_cxl_endpoint() conditional, just split the two cases. Rather than duplicate the "Couldn't enumerate decoders" error message take the opportunity to improve the error messages in devm_cxl_enumerate_decoders(). Reported-by:
-
Dan Williams authored
Add help text and a label so the CXL_REGION config option can be toggled. This is mainly to enable compile testing without region support. Reviewed-by:
-
Dan Williams authored
In support of the CXL subsystem's use of 'struct range' to track decode address ranges, add a common range_contains() implementation with identical semantics as resource_contains(); The existing 'range_contains()' in lib/stackinit_kunit.c is namespaced with a 'stackinit_' prefix. Cc: Kees Cook <keescook@chromium.org> Reviewed-by:
-
Dan Williams authored
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:
-
Dan Williams authored
Jonathan noticed that the target list setup is not unwound completely upon error. Undo all the setup in the 'err_decrement:' exit path. Fixes: 27b3f8d1 ("cxl/region: Program target lists") Reported-by:
-
Dan Williams authored
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:
-
Dan Williams authored
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:
-
Dan Williams authored
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:
-
Dan Williams authored
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:
-
Dan Williams authored
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:
-
Dan Williams authored
Prior to Linus deciding that the kernel that following v5.19 would be v6.0, the CXL ABI documentation already referenced v5.20. In preparation for updating these entries update the kernel version to v6.0. Reviewed-by:
Davidlohr Bueso <dave@stgolabs.net> Reviewed-by:
-
Dan Williams authored
Testing of ram region support [1], stimulates a long standing bug in cxl_detach_ep() where some cxl_ep_remove() cleanup is skipped due to inability to walk ports after dports have been unregistered. That results in a failure to re-register a memdev after the port is re-enabled leading to a crash like the following: cxl_port_setup_targets: cxl region4: cxl_host_bridge.0:port4 iw: 1 ig: 256 general protection fault, ... [..] RIP: 0010:cxl_region_setup_targets+0x897/0x9e0 [cxl_core] dev_name at include/linux/device.h:700 (inlined by) cxl_port_setup_targets at drivers/cxl/core/region.c:1155 (inlined by) cxl_region_setup_targets at drivers/cxl/core/region.c:1249 [..] Call Trace: <TASK> attach_target+0x39a/0x760 [cxl_core] ? __mutex_unlock_slowpath+0x3a/0x290 cxl_add_to_region+0xb8/0x340 [cxl_core] ? lockdep_hardirqs_on+0x7d/0x100 discover_region+0x4b/0x80 [cxl_port] ? __pfx_discover_region+0x10/0x10 [cxl_port] device_for_each_child+0x58/0x90 cxl_port_probe+0x10e/0x130 [cxl_port] cxl_bus_probe+0x17/0x50 [cxl_core] Change the port ancestry walk to be by depth rather than by dport. This ensures that even if a port has unregistered its dports a deferred memdev cleanup will still be able to cleanup the memdev's interest in that port. The parent_port->dev.driver check is only needed for determining if the bottom up removal beat the top-down removal, but cxl_ep_remove() can always proceed given the port is pinned. That is, the two sources of cxl_ep_remove() are in cxl_detach_ep() and cxl_port_release(), and cxl_port_release() can not run if cxl_detach_ep() holds a reference. Fixes: 2703c16c ("cxl/core/port: Add switch port enumeration") Link: http://lore.kernel.org/r/167564534874.847146.5222419648551436750.stgit@dwillia2-xfh.jf.intel.com [1] Reviewed-by:
-
- 09 Feb, 2023 1 commit
-
-
Davidlohr Bueso authored
For ID allocations we want 0-(max-1), ie: smatch complains: error: Calling ida_alloc_range() with a 'max' argument which is a power of 2. -1 missing? Correct this and also replace the call to use the max() flavor instead. Signed-off-by:
-
- 07 Feb, 2023 4 commits
-
-
Dan Williams authored
Add the CXL event and interrupt support for the v6.3 update.
-
Dan Williams authored
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.
-
Dan Williams authored
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: e4f6dfa9 ("cxl/region: Fix 'distance' calculation with passthrough ports") Cc: <stable@vger.kernel.org> Reviewed-by:
-
Fan Ni authored
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: 176baefb ("cxl/hdm: Commit decoder state to hardware") Cc: <stable@vger.kernel.org> Signed-off-by:
-
- 30 Jan, 2023 3 commits
-
-
Dan Williams authored
The IRQ core expects that users of the default hardirq handler specify IRQF_ONESHOT to keep interrupts disabled until the threaded handler runs. That meets the CXL driver's expectations since it is an edge triggered MSI and this flag would have been passed by default using pci_request_irq() instead of devm_request_threaded_irq(). Fixes: a49aa814 ("cxl/mem: Wire up event interrupts") Reported-by:
kernel test robot <lkp@intel.com> Reported-by:
Julia Lawall <julia.lawall@lip6.fr> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: Ira Weiny <ira.weiny@intel.com> Signed-off-by:
-
Jonathan Cameron authored
CXL r3.0 section 8.2.9.4.2 "Set Timestamp" recommends that the host sets the timestamp after every Conventional or CXL Reset to ensure accurate timestamps. This should include on initial boot up. The time base that is being set is used by a device for the poison list overflow timestamp and all event timestamps. Note that the command is optional and if not supported and the device cannot return accurate timestamps it will fill the fields in with an appropriate marker (see the specification description of each timestamp). Signed-off-by:
-
Jonathan Cameron authored
Kernel-doc should be complete, so add documentation for the status parameter. Signed-off-by:
-
- 29 Jan, 2023 6 commits
-
-
Linus Torvalds authored
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull irq fix from Borislav Petkov: - Cleanup the firmware node for the new IRQ MSI domain properly, to avoid leaking memory * tag 'irq_urgent_for_v6.2_rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: genirq/msi: Free the fwnode created by msi_create_device_irq_domain()
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 fixes from Borislav Petkov: - Start checking for -mindirect-branch-cs-prefix clang support too now that LLVM 16 will support it - Fix a NULL ptr deref when suspending with Xen PV - Have a SEV-SNP guest check explicitly for features enabled by the hypervisor and fail gracefully if some are unsupported by the guest instead of failing in a non-obvious and hard-to-debug way - Fix a MSI descriptor leakage under Xen - Mark Xen's MSI domain as supporting MSI-X - Prevent legacy PIC interrupts from being resent in software by marking them level triggered, as they should be, which lead to a NULL ptr deref * tag 'x86_urgent_for_v6.2_rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/build: Move '-mindirect-branch-cs-prefix' out of GCC-only block acpi: Fix suspend with Xen PV x86/sev: Add SEV-SNP guest feature negotiation support x86/pci/xen: Fixup fallout from the PCI/MSI overhaul x86/pci/xen: Set MSI_FLAG_PCI_MSIX support in Xen MSI domain x86/i8259: Mark legacy PIC interrupts with IRQ_LEVEL
-
git://git.kernel.org/pub/scm/linux/kernel/git/dtor/inputLinus Torvalds authored
Pull input fixes from Dmitry Torokhov: - touchpads on HP 15-* laptops switched back to PS/2 emulation mode - a quirk for Clevo PCX0DX/TUXEDO XP1511 to make sure keyboard is responding after resume * tag 'input-for-v6.2-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/dtor/input: Input: i8042 - add Clevo PCX0DX to i8042 quirk table Revert "Input: synaptics - switch touchpad on HP Laptop 15-da3001TU to RMI mode"
-
git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxlLinus Torvalds authored
Pull cxl fixes from Dan Williams: "A couple of fixes for bugs introduced during the merge window. One is a regression, the other was a bug in the CXL AER handler: - Fix a crash regression due to module load order of cxl_pmem.ko - Fix wrong register offset read in CXL AER handling path" * tag 'cxl-fixes-for-6.2-rc6' of git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl: cxl/pmem: Fix nvdimm unregistration when cxl_pmem driver is absent cxl: fix cxl_report_and_clear() RAS UE addr mis-assignment
-
Vlastimil Babka authored
This reverts commit 7efc3b72. We have got openSUSE reports (Link 1) for 6.1 kernel with khugepaged stalling CPU for long periods of time. Investigation of tracepoint data shows that compaction is stuck in repeating fast_find_migrateblock() based migrate page isolation, and then fails to migrate all isolated pages. Commit 7efc3b72 ("mm/compaction: fix set skip in fast_find_migrateblock") was suspected as it was merged in 6.1 and in theory can indeed remove a termination condition for fast_find_migrateblock() under certain conditions, as it removes a place that always marks a scanned pageblock from being re-scanned. There are other such places, but those can be skipped under certain conditions, which seems to match the tracepoint data. Testing of revert also appears to have resolved the issue, thus revert the commit until a more robust solution for the original problem is developed. It's also likely this will fix qemu stalls with 6.1 kernel reported in Link 2, but that is not yet confirmed. Link: https://bugzilla.suse.com/show_bug.cgi?id=1206848 Link: https://lore.kernel.org/kvm/b8017e09-f336-3035-8344-c549086c2340@kernel.org/ Link: https://lore.kernel.org/lkml/20230125134434.18017-1-mgorman@techsingularity.net/ Fixes: 7efc3b72 ("mm/compaction: fix set skip in fast_find_migrateblock") Cc: <stable@vger.kernel.org> Tested-by:
Pedro Falcato <pedro.falcato@gmail.com> Acked-by:
Mel Gorman <mgorman@techsingularity.net> Signed-off-by:
Vlastimil Babka <vbabka@suse.cz> Signed-off-by:
Linus Torvalds <torvalds@linux-foundation.org>
-
