> -----Original Message----- > From: Linux-nvdimm [mailto:linux-nvdimm-boun...@lists.01.org] On > Behalf Of Dave Jiang > Sent: Monday, December 12, 2016 6:15 PM > Subject: [PATCH 1/3] dev-dax: add support to display badblocks in > sysfs for dev-dax > > Adding support to show badblocks in the pmem region that's provided > by the poison_list. This should show up in > /sys/class/dax/daxN.N/badblocks as read only. Currently we only > support > a single resource and we do not support badblocks for seeds. > Additional > support code will be implemented to support. > ... > diff --git a/drivers/dax/pmem.c b/drivers/dax/pmem.c ... > +static ssize_t dax_pmem_badblocks_show(struct device *dev, > + struct device_attribute *attr, char *page) > +{ > + struct dax_pmem *dax_pmem = > + (struct dax_pmem *)dax_dev_get_private(dev); > + > + return badblocks_show(&dax_pmem->bb, page, 0); > +}
Do /dev/dax devices have a "block size" to explain the units used in the sysfs files? If so, where is it reported? This directory has the total size in bytes (here, 8 GiB - 2 MiB): /sys/class/dax/dax0.0/dev:242:0 /sys/class/dax/dax0.0/size:8587837440 /sys/class/dax/dax0.0/uevent:MAJOR=242 /sys/class/dax/dax0.0/uevent:MINOR=0 /sys/class/dax/dax0.0/uevent:DEVNAME=dax0.0 In contrast, /dev/pmem devices have a full "queue" directory, including logical_block_size. /sys/class/block/pmem0/device/devtype:nd_namespace_io /sys/class/block/pmem0/device/force_raw:0 /sys/class/block/pmem0/device/modalias:nd:t4 /sys/class/block/pmem0/device/mode:raw /sys/class/block/pmem0/device/nstype:4 /sys/class/block/pmem0/device/numa_node:0 /sys/class/block/pmem0/device/resource:0x880000000 /sys/class/block/pmem0/device/size:8589934592 /sys/class/block/pmem0/device/uevent:DEVTYPE=nd_namespace_io /sys/class/block/pmem0/device/uevent:DRIVER=nd_pmem /sys/class/block/pmem0/device/uevent:MODALIAS=nd:t4 /sys/class/block/pmem0/queue/add_random:0 /sys/class/block/pmem0/queue/dax:1 /sys/class/block/pmem0/queue/discard_granularity:0 /sys/class/block/pmem0/queue/discard_max_bytes:0 /sys/class/block/pmem0/queue/discard_max_hw_bytes:0 /sys/class/block/pmem0/queue/discard_zeroes_data:0 /sys/class/block/pmem0/queue/hw_sector_size:512 /sys/class/block/pmem0/queue/io_poll:0 /sys/class/block/pmem0/queue/iostats:0 /sys/class/block/pmem0/queue/logical_block_size:512 /sys/class/block/pmem0/queue/max_hw_sectors_kb:2147483647 /sys/class/block/pmem0/queue/max_integrity_segments:0 /sys/class/block/pmem0/queue/max_sectors_kb:1280 /sys/class/block/pmem0/queue/max_segments:128 /sys/class/block/pmem0/queue/max_segment_size:65536 /sys/class/block/pmem0/queue/minimum_io_size:4096 /sys/class/block/pmem0/queue/nomerges:0 /sys/class/block/pmem0/queue/nr_requests:128 /sys/class/block/pmem0/queue/optimal_io_size:0 /sys/class/block/pmem0/queue/physical_block_size:4096 /sys/class/block/pmem0/queue/read_ahead_kb:128 /sys/class/block/pmem0/queue/rotational:0 /sys/class/block/pmem0/queue/rq_affinity:0 /sys/class/block/pmem0/queue/scheduler:none /sys/class/block/pmem0/queue/write_cache:write back /sys/class/block/pmem0/queue/write_same_max_bytes:0 --- Robert Elliott, HPE Persistent Memory _______________________________________________ Linux-nvdimm mailing list Linux-nvdimm@lists.01.org https://lists.01.org/mailman/listinfo/linux-nvdimm
