On Sun, 11 Oct 2026 07:54:53 +0900 Anisa Su <[email protected]> wrote:
> From: Ira Weiny <[email protected]> > > Devices which support Dynamic Capacity (DC) are configured via mailbox > commands. CXL r4.0 section 9.13.3 describes the use of the Get DC > Configuration command in order to obtain DCD partition characteristics. > > Implement the DC mailbox commands as specified in CXL 4.0 section > 8.2.10.9.9 (opcodes 48XXh) to read and store the DCD configuration > information. Disable DCD if an invalid configuration is found. > > Initial enablement for Linux only supports one dynamic capacity > partition. Read all the partitions the device reports but validate > only the first, and configure it as 'dynamic ram'. > > The specification requires that volatile capacity starts at DPA 0 and > pmem starts at the DPA immediately following it, but DC partitions only > need to be 256MB aligned according to CXL r4.0 section 8.2.10.9.9.1 > Table 8-347. So a device could leave a gap between ram/pmem (static) > capacity and its first DC partition, or between one DC partition and > the next. > > However, Linux chooses not to support gaps between static and dynamic > capacity, and requires the first DC partition to begin at the DPA > immediately following static capacity. > > Based on an original patch by Navneet Singh. > > Signed-off-by: Ira Weiny <[email protected]> > Co-developed-by: Anisa Su <[email protected]> > Signed-off-by: Anisa Su <[email protected]> > Tested-by: Wonjae Lee <[email protected]> > Tested-by: Junhee Park <[email protected]> > Tested-by: Heesoo Kim <[email protected]> > A couple of things inline. Only one that matters I think is the cleanup.h fragility / readability thing about constructors and destructors needing to be near each other. That bit of guidance came from a specific (and I think correct) comment from Linus that he really wanted them together and if that meant declaring the variables inline rather than in old school c style, then so be it. That might be in category of things Dave could cleanup up whilst applying if all else is ready but that's up to him. Jonathan > diff --git a/drivers/cxl/core/mbox.c b/drivers/cxl/core/mbox.c > index 199bb986d674..72af23ba288e 100644 > --- a/drivers/cxl/core/mbox.c > +++ b/drivers/cxl/core/mbox.c > +/* Returns the number of partitions in dc_resp or -ERRNO */ > +static int cxl_get_dc_config(struct cxl_mailbox *mbox, u8 start_partition, > + u8 partition_count, > + struct cxl_mbox_get_dc_config_out *dc_resp, > + size_t dc_resp_size) > +{ > + struct cxl_mbox_get_dc_config_in get_dc = (struct > cxl_mbox_get_dc_config_in) { > + .partition_count = partition_count, > + .start_partition_index = start_partition, > + }; > + struct cxl_mbox_cmd mbox_cmd = (struct cxl_mbox_cmd) { > + .opcode = CXL_MBOX_OP_GET_DC_CONFIG, > + .payload_in = &get_dc, > + .size_in = sizeof(get_dc), > + .size_out = dc_resp_size, > + .payload_out = dc_resp, > + /* The device must return at least the fixed header */ > + .min_out = sizeof(*dc_resp), > + }; > + size_t expected_sz; > + int rc; > + > + rc = cxl_internal_send_cmd(mbox, &mbox_cmd); > + if (rc < 0) > + return rc; > + > + if (dc_resp->partitions_returned > partition_count) { > + dev_err(mbox->host, "Device returned %u partitions, requested > %u\n", > + dc_resp->partitions_returned, partition_count); > + return -EIO; > + } > + > + /* The trailing extent/tag counts (CXL r4.0 Table 8-346) are not read */ > + expected_sz = struct_size(dc_resp, partition, > + dc_resp->partitions_returned); > + > + if (mbox_cmd.size_out < expected_sz) { > + dev_err(mbox->host, > + "Payload size %zu less than expected %zu for %u > partitions\n", > + mbox_cmd.size_out, > + expected_sz, > + dc_resp->partitions_returned); Not important, but if you do end up respinning, maybe don't have one parameter per line. To me it doesn't really help with readability here. > + return -EIO; > + } > + > +/** > + * cxl_dev_dc_identify() - Reads the dynamic capacity information from the > + * device. > + * @mbox: Mailbox to query > + * @dc_info: The dynamic partition information to return > + * > + * Read every partition the device reports, but validate only the first: > + * Linux maps partition 0 and nothing else, so a defect in capacity the > + * driver never touches is not a reason to refuse the device dynamic > + * capacity. The remaining partitions are read but not recorded. > + * > + * Return: 0 if identify was executed successfully, -ERRNO on error. > + * On error, @dc_info is left unchanged. > + */ > +int cxl_dev_dc_identify(struct cxl_mailbox *mbox, > + struct cxl_dc_partition_info *dc_info) > +{ > + struct cxl_dc_partition_info partitions[CXL_MAX_DC_PARTITIONS] = { }; > + struct cxl_mbox_get_dc_config_out *dc_resp __free(kfree) = NULL; See the cleanup.h guidance. This needs to move. > + struct device *dev = mbox->host; > + u8 avail_partitions = 0; > + u8 start_partition; > + u8 num_partitions; > + u8 partition_count; > + size_t dc_resp_size; > + > + /* > + * Bound requested number of partitions by mailbox payload size. The > + * 256 byte spec minimum, verified in cxl_pci_setup_mailbox(), keeps > + * the subtraction below from underflowing. > + */ > + partition_count = min(CXL_MAX_DC_PARTITIONS, > + (mbox->payload_size - sizeof(*dc_resp) - > + sizeof(struct cxl_mbox_get_dc_config_tail)) / > + sizeof(struct cxl_dc_partition)); > + dc_resp_size = struct_size(dc_resp, partition, partition_count) + > + sizeof(struct cxl_mbox_get_dc_config_tail); > + struct cxl_mbox_get_dc_config_out *dc_resp __free(kfree) = kmalloc(dc_resp_size, GFP_KERNEL); if (!dc_resp) Maybe if everything else is fine, Dave can tweak this whilst applying. star> + if (!dc_resp) > + return -ENOMEM; > + > + start_partition = 0; > + num_partitions = 0; > + do { > + int returned, rc; > + > + returned = cxl_get_dc_config(mbox, start_partition, > + partition_count, dc_resp, > + dc_resp_size); > + if (returned < 0) { > + dev_err(dev, "Failed to get DC config: %d\n", returned); > + return returned; > + } > + > + if (returned == 0) { > + dev_err(dev, > + "Device reported %u partitions available but > returned none at index %u\n", > + dc_resp->avail_partition_count, > start_partition); > + return -EIO; > + } > + > + /* A DCD reports at least one partition (CXL r4.0 Table 8-346) > */ > + if (dc_resp->avail_partition_count == 0) { > + dev_err(dev, "Device reported no DC partitions\n"); > + return -EIO; > + } > + > + avail_partitions = min(dc_resp->avail_partition_count, > + CXL_MAX_DC_PARTITIONS); > + > + /* > + * Only partition 0 is mapped, so it is the only one checked. > + * The rest are read to complete the enumeration and dropped. > + */ > + if (start_partition == 0) { > + rc = cxl_dc_check(dev, &partitions[0], > + &dc_resp->partition[0]); > + if (rc) > + return rc; > + } > + > + num_partitions += returned; > + start_partition = num_partitions; I may be missing something subtle, but do start_partition and num_partitions ever differ in value? Far as I can spot they are both only updated here. If they change in later patches and this is churn reduction ignore this comment! > + > + } while (num_partitions < avail_partitions); > + > + /* Linux only supports the 1st partition; nothing to do if it is > unavailable */ > + if (partitions[0].size == 0) > + return -ENODEV; > + > + *dc_info = partitions[0]; > + dev_dbg(dev, "Returning partition 0 %#llx size %#llx\n", > + dc_info->start, dc_info->size); > + > + return 0; > +} > +EXPORT_SYMBOL_NS_GPL(cxl_dev_dc_identify, "CXL");
