--- /dev/null
+======
+dm-ima
+======
+
+For a given system, various external services/infrastructure tools
+(including the attestation service) interact with it - both during the
+setup and during rest of the system run-time. They share sensitive data
+and/or execute critical workload on that system. The external services
+may want to verify the current run-time state of the relevant kernel
+subsystems before fully trusting the system with business-critical
+data/workload.
+
+Device mapper plays a critical role on a given system by providing
+various important functionalities to the block devices using various
+target types like crypt, verity, integrity etc. Each of these target
+types’ functionalities can be configured with various attributes.
+The attributes chosen to configure these target types can significantly
+impact the security profile of the block device, and in-turn, of the
+system itself. For instance, the type of encryption algorithm and the
+key size determines the strength of encryption for a given block device.
+
+Therefore, verifying the current state of various block devices as well
+as their various target attributes is crucial for external services before
+fully trusting the system with business-critical data/workload.
+
+IMA kernel subsystem provides the necessary functionality for
+device mapper to measure the state and configuration of
+various block devices -
+
+- by device mapper itself, from within the kernel,
+- in a tamper resistant way,
+- and re-measured - triggered on state/configuration change.
+
+Setting the IMA Policy:
+=======================
+For IMA to measure the data on a given system, the IMA policy on the
+system needs to be updated to have following line, and the system needs
+to be restarted for the measurements to take effect.
+
+::
+
+ /etc/ima/ima-policy
+ measure func=CRITICAL_DATA label=device-mapper template=ima-buf
+
+The measurements will be reflected in the IMA logs, which are located at:
+
+::
+
+ /sys/kernel/security/integrity/ima/ascii_runtime_measurements
+ /sys/kernel/security/integrity/ima/binary_runtime_measurements
+
+Then IMA ASCII measurement log has the following format:
+
+::
+
+ <PCR> <TEMPLATE_DATA_DIGEST> <TEMPLATE_NAME> <TEMPLATE_DATA>
+
+ PCR := Platform Configuration Register, in which the values are registered.
+ This is applicable if TPM chip is in use.
+
+ TEMPLATE_DATA_DIGEST := Template data digest of the IMA record.
+ TEMPLATE_NAME := Template name that registered the integrity value (e.g. ima-buf).
+
+ TEMPLATE_DATA := <ALG> ":" <EVENT_DIGEST> <EVENT_NAME> <EVENT_DATA>
+ It contains data for the specific event to be measured,
+ in a given template data format.
+
+ ALG := Algorithm to compute event digest
+ EVENT_DIGEST := Digest of the event data
+ EVENT_NAME := Description of the event (e.g. 'dm_table_load').
+ EVENT_DATA := The event data to be measured.
+
+|
+
+| *NOTE #1:*
+| The DM target data measured by IMA subsystem can alternatively
+ be queried from userspace by setting DM_IMA_MEASUREMENT_FLAG with
+ DM_TABLE_STATUS_CMD.
+
+|
+
+| *NOTE #2:*
+| The Kernel configuration CONFIG_IMA_DISABLE_HTABLE allows measurement of duplicate records.
+| To support recording duplicate IMA events in the IMA log, the Kernel needs to be configured with
+ CONFIG_IMA_DISABLE_HTABLE=y.
+
+Supported Device States:
+========================
+Following device state changes will trigger IMA measurements:
+
+ 1. Table load
+ #. Device resume
+ #. Device remove
+ #. Table clear
+ #. Device rename
+
+1. Table load:
+---------------
+When a new table is loaded in a device's inactive table slot,
+the device information and target specific details from the
+targets in the table are measured.
+
+The IMA measurement log has the following format for 'dm_table_load':
+
+::
+
+ EVENT_NAME := "dm_table_load"
+ EVENT_DATA := <dm_version_str> ";" <device_metadata> ";" <table_load_data>
+
+ dm_version_str := "dm_version=" <N> "." <N> "." <N>
+ Same as Device Mapper driver version.
+ device_metadata := <device_name> "," <device_uuid> "," <device_major> "," <device_minor> ","
+ <minor_count> "," <num_device_targets> ";"
+
+ device_name := "name=" <dm-device-name>
+ device_uuid := "uuid=" <dm-device-uuid>
+ device_major := "major=" <N>
+ device_minor := "minor=" <N>
+ minor_count := "minor_count=" <N>
+ num_device_targets := "num_targets=" <N>
+ dm-device-name := Name of the device. If it contains special characters like '\', ',', ';',
+ they are prefixed with '\'.
+ dm-device-uuid := UUID of the device. If it contains special characters like '\', ',', ';',
+ they are prefixed with '\'.
+
+ table_load_data := <target_data>
+ Represents the data (as name=value pairs) from various targets in the table,
+ which is being loaded into the DM device's inactive table slot.
+ target_data := <target_data_row> | <target_data><target_data_row>
+
+ target_data_row := <target_index> "," <target_begin> "," <target_len> "," <target_name> ","
+ <target_version> "," <target_attributes> ";"
+ target_index := "target_index=" <N>
+ Represents nth target in the table (from 0 to N-1 targets specified in <num_device_targets>)
+ If all the data for N targets doesn't fit in the given buffer - then the data that fits
+ in the buffer (say from target 0 to x) is measured in a given IMA event.
+ The remaining data from targets x+1 to N-1 is measured in the subsequent IMA events,
+ with the same format as that of 'dm_table_load'
+ i.e. <dm_version_str> ";" <device_metadata> ";" <table_load_data>.
+
+ target_begin := "target_begin=" <N>
+ target_len := "target_len=" <N>
+ target_name := Name of the target. 'linear', 'crypt', 'integrity' etc.
+ The targets that are supported for IMA measurements are documented below in the
+ 'Supported targets' section.
+ target_version := "target_version=" <N> "." <N> "." <N>
+ target_attributes := Data containing comma separated list of name=value pairs of target specific attributes.
+
+ For instance, if a linear device is created with the following table entries,
+ # dmsetup create linear1
+ 0 2 linear /dev/loop0 512
+ 2 2 linear /dev/loop0 512
+ 4 2 linear /dev/loop0 512
+ 6 2 linear /dev/loop0 512
+
+ Then IMA ASCII measurement log will have the following entry:
+ (converted from ASCII to text for readability)
+
+ 10 a8c5ff755561c7a28146389d1514c318592af49a ima-buf sha256:4d73481ecce5eadba8ab084640d85bb9ca899af4d0a122989252a76efadc5b72
+ dm_table_load
+ dm_version=4.45.0;
+ name=linear1,uuid=,major=253,minor=0,minor_count=1,num_targets=4;
+ target_index=0,target_begin=0,target_len=2,target_name=linear,target_version=1.4.0,device_name=7:0,start=512;
+ target_index=1,target_begin=2,target_len=2,target_name=linear,target_version=1.4.0,device_name=7:0,start=512;
+ target_index=2,target_begin=4,target_len=2,target_name=linear,target_version=1.4.0,device_name=7:0,start=512;
+ target_index=3,target_begin=6,target_len=2,target_name=linear,target_version=1.4.0,device_name=7:0,start=512;
+
+2. Device resume:
+------------------
+When a suspended device is resumed, the device information and the hash of the
+data from previous load of an active table are measured.
+
+The IMA measurement log has the following format for 'dm_device_resume':
+
+::
+
+ EVENT_NAME := "dm_device_resume"
+ EVENT_DATA := <dm_version_str> ";" <device_metadata> ";" <active_table_hash> ";" <current_device_capacity> ";"
+
+ dm_version_str := As described in the 'Table load' section above.
+ device_metadata := As described in the 'Table load' section above.
+ active_table_hash := "active_table_hash=" <table_hash_alg> ":" <table_hash>
+ Rerpresents the hash of the IMA data being measured for the
+ active table for the device.
+ table_hash_alg := Algorithm used to compute the hash.
+ table_hash := Hash of the (<dm_version_str> ";" <device_metadata> ";" <table_load_data> ";")
+ as described in the 'dm_table_load' above.
+ Note: If the table_load data spans across multiple IMA 'dm_table_load'
+ events for a given device, the hash is computed combining all the event data
+ i.e. (<dm_version_str> ";" <device_metadata> ";" <table_load_data> ";")
+ across all those events.
+ current_device_capacity := "current_device_capacity=" <N>
+
+ For instance, if a linear device is resumed with the following command,
+ #dmsetup resume linear1
+
+ then IMA ASCII measurement log will have an entry with:
+ (converted from ASCII to text for readability)
+
+ 10 56c00cc062ffc24ccd9ac2d67d194af3282b934e ima-buf sha256:e7d12c03b958b4e0e53e7363a06376be88d98a1ac191fdbd3baf5e4b77f329b6
+ dm_device_resume
+ dm_version=4.45.0;
+ name=linear1,uuid=,major=253,minor=0,minor_count=1,num_targets=4;
+ active_table_hash=sha256:4d73481ecce5eadba8ab084640d85bb9ca899af4d0a122989252a76efadc5b72;current_device_capacity=8;
+
+3. Device remove:
+------------------
+When a device is removed, the device information and a sha256 hash of the
+data from an active and inactive table are measured.
+
+The IMA measurement log has the following format for 'dm_device_remove':
+
+::
+
+ EVENT_NAME := "dm_device_remove"
+ EVENT_DATA := <dm_version_str> ";" <device_active_metadata> ";" <device_inactive_metadata> ";"
+ <active_table_hash> "," <inactive_table_hash> "," <remove_all> ";" <current_device_capacity> ";"
+
+ dm_version_str := As described in the 'Table load' section above.
+ device_active_metadata := Device metadata that reflects the currently loaded active table.
+ The format is same as 'device_metadata' described in the 'Table load' section above.
+ device_inactive_metadata := Device metadata that reflects the inactive table.
+ The format is same as 'device_metadata' described in the 'Table load' section above.
+ active_table_hash := Hash of the currently loaded active table.
+ The format is same as 'active_table_hash' described in the 'Device resume' section above.
+ inactive_table_hash := Hash of the inactive table.
+ The format is same as 'active_table_hash' described in the 'Device resume' section above.
+ remove_all := "remove_all=" <yes_no>
+ yes_no := "y" | "n"
+ current_device_capacity := "current_device_capacity=" <N>
+
+ For instance, if a linear device is removed with the following command,
+ #dmsetup remove l1
+
+ then IMA ASCII measurement log will have the following entry:
+ (converted from ASCII to text for readability)
+
+ 10 790e830a3a7a31590824ac0642b3b31c2d0e8b38 ima-buf sha256:ab9f3c959367a8f5d4403d6ce9c3627dadfa8f9f0e7ec7899299782388de3840
+ dm_device_remove
+ dm_version=4.45.0;
+ device_active_metadata=name=l1,uuid=,major=253,minor=2,minor_count=1,num_targets=2;
+ device_inactive_metadata=name=l1,uuid=,major=253,minor=2,minor_count=1,num_targets=1;
+ active_table_hash=sha256:4a7e62efaebfc86af755831998b7db6f59b60d23c9534fb16a4455907957953a,
+ inactive_table_hash=sha256:9d79c175bc2302d55a183e8f50ad4bafd60f7692fd6249e5fd213e2464384b86,remove_all=n;
+ current_device_capacity=2048;
+
+4. Table clear:
+----------------
+When an inactive table is cleared from the device, the device information and a sha256 hash of the
+data from an inactive table are measured.
+
+The IMA measurement log has the following format for 'dm_table_clear':
+
+::
+
+ EVENT_NAME := "dm_table_clear"
+ EVENT_DATA := <dm_version_str> ";" <device_inactive_metadata> ";" <inactive_table_hash> ";" <current_device_capacity> ";"
+
+ dm_version_str := As described in the 'Table load' section above.
+ device_inactive_metadata := Device metadata that was captured during the load time inactive table being cleared.
+ The format is same as 'device_metadata' described in the 'Table load' section above.
+ inactive_table_hash := Hash of the inactive table being cleared from the device.
+ The format is same as 'active_table_hash' described in the 'Device resume' section above.
+ current_device_capacity := "current_device_capacity=" <N>
+
+ For instance, if a linear device's inactive table is cleared,
+ #dmsetup clear l1
+
+ then IMA ASCII measurement log will have an entry with:
+ (converted from ASCII to text for readability)
+
+ 10 77d347408f557f68f0041acb0072946bb2367fe5 ima-buf sha256:42f9ca22163fdfa548e6229dece2959bc5ce295c681644240035827ada0e1db5
+ dm_table_clear
+ dm_version=4.45.0;
+ name=l1,uuid=,major=253,minor=2,minor_count=1,num_targets=1;
+ inactive_table_hash=sha256:75c0dc347063bf474d28a9907037eba060bfe39d8847fc0646d75e149045d545;current_device_capacity=1024;
+
+5. Device rename:
+------------------
+When an device's NAME or UUID is changed, the device information and the new NAME and UUID
+are measured.
+
+The IMA measurement log has the following format for 'dm_device_rename':
+
+::
+
+ EVENT_NAME := "dm_device_rename"
+ EVENT_DATA := <dm_version_str> ";" <device_active_metadata> ";" <new_device_name> "," <new_device_uuid> ";" <current_device_capacity> ";"
+
+ dm_version_str := As described in the 'Table load' section above.
+ device_active_metadata := Device metadata that reflects the currently loaded active table.
+ The format is same as 'device_metadata' described in the 'Table load' section above.
+ new_device_name := "new_name=" <dm-device-name>
+ dm-device-name := Same as <dm-device-name> described in 'Table load' section above
+ new_device_uuid := "new_uuid=" <dm-device-uuid>
+ dm-device-uuid := Same as <dm-device-uuid> described in 'Table load' section above
+ current_device_capacity := "current_device_capacity=" <N>
+
+ E.g 1: if a linear device's name is changed with the following command,
+ #dmsetup rename linear1 --setuuid 1234-5678
+
+ then IMA ASCII measurement log will have an entry with:
+ (converted from ASCII to text for readability)
+
+ 10 8b0423209b4c66ac1523f4c9848c9b51ee332f48 ima-buf sha256:6847b7258134189531db593e9230b257c84f04038b5a18fd2e1473860e0569ac
+ dm_device_rename
+ dm_version=4.45.0;
+ name=linear1,uuid=,major=253,minor=2,minor_count=1,num_targets=1;new_name=linear1,new_uuid=1234-5678;
+ current_device_capacity=1024;
+
+ E.g 2: if a linear device's name is changed with the following command,
+ # dmsetup rename linear1 linear=2
+
+ then IMA ASCII measurement log will have an entry with:
+ (converted from ASCII to text for readability)
+
+ 10 bef70476b99c2bdf7136fae033aa8627da1bf76f ima-buf sha256:8c6f9f53b9ef9dc8f92a2f2cca8910e622543d0f0d37d484870cb16b95111402
+ dm_device_rename
+ dm_version=4.45.0;
+ name=linear1,uuid=1234-5678,major=253,minor=2,minor_count=1,num_targets=1;
+ new_name=linear\=2,new_uuid=1234-5678;
+ current_device_capacity=1024;
+
+Supported targets:
+==================
+
+Following targets are supported to measure their data using IMA:
+
+ 1. cache
+ #. crypt
+ #. integrity
+ #. linear
+ #. mirror
+ #. multipath
+ #. raid
+ #. snapshot
+ #. striped
+ #. verity
+
+1. cache
+---------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'cache' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <metadata_mode> "," <cache_metadata_device> ","
+ <cache_device> "," <cache_origin_device> "," <writethrough> "," <writeback> ","
+ <passthrough> "," <no_discard_passdown> ";"
+
+ target_name := "target_name=cache"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ metadata_mode := "metadata_mode=" <cache_metadata_mode>
+ cache_metadata_mode := "fail" | "ro" | "rw"
+ cache_device := "cache_device=" <cache_device_name_string>
+ cache_origin_device := "cache_origin_device=" <cache_origin_device_string>
+ writethrough := "writethrough=" <yes_no>
+ writeback := "writeback=" <yes_no>
+ passthrough := "passthrough=" <yes_no>
+ no_discard_passdown := "no_discard_passdown=" <yes_no>
+ yes_no := "y" | "n"
+
+ E.g.
+ When a 'cache' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'cache' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;name=cache1,uuid=cache_uuid,major=253,minor=2,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=28672,target_name=cache,target_version=2.2.0,metadata_mode=rw,
+ cache_metadata_device=253:4,cache_device=253:3,cache_origin_device=253:5,writethrough=y,writeback=n,
+ passthrough=n,metadata2=y,no_discard_passdown=n;
+
+
+2. crypt
+---------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'crypt' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <allow_discards> "," <same_cpu_crypt> ","
+ <submit_from_crypt_cpus> "," <no_read_workqueue> "," <no_write_workqueue> ","
+ <iv_large_sectors> "," <iv_large_sectors> "," [<integrity_tag_size> ","] [<cipher_auth> ","]
+ [<sector_size> ","] [<cipher_string> ","] <key_size> "," <key_parts> ","
+ <key_extra_size> "," <key_mac_size> ";"
+
+ target_name := "target_name=crypt"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ allow_discards := "allow_discards=" <yes_no>
+ same_cpu_crypt := "same_cpu_crypt=" <yes_no>
+ submit_from_crypt_cpus := "submit_from_crypt_cpus=" <yes_no>
+ no_read_workqueue := "no_read_workqueue=" <yes_no>
+ no_write_workqueue := "no_write_workqueue=" <yes_no>
+ iv_large_sectors := "iv_large_sectors=" <yes_no>
+ integrity_tag_size := "integrity_tag_size=" <N>
+ cipher_auth := "cipher_auth=" <string>
+ sector_size := "sector_size=" <N>
+ cipher_string := "cipher_string="
+ key_size := "key_size=" <N>
+ key_parts := "key_parts=" <N>
+ key_extra_size := "key_extra_size=" <N>
+ key_mac_size := "key_mac_size=" <N>
+ yes_no := "y" | "n"
+
+ E.g.
+ When a 'crypt' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'crypt' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=crypt1,uuid=crypt_uuid1,major=253,minor=0,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=1953125,target_name=crypt,target_version=1.23.0,
+ allow_discards=y,same_cpu=n,submit_from_crypt_cpus=n,no_read_workqueue=n,no_write_workqueue=n,
+ iv_large_sectors=n,cipher_string=aes-xts-plain64,key_size=32,key_parts=1,key_extra_size=0,key_mac_size=0;
+
+3. integrity
+-------------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'integrity' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <dev_name> "," <start>
+ <tag_size> "," <mode> "," [<meta_device> ","] [<block_size> ","] <recalculate> ","
+ <allow_discards> "," <fix_padding> "," <fix_hmac> "," <legacy_recalculate> ","
+ <journal_sectors> "," <interleave_sectors> "," <buffer_sectors> ";"
+
+ target_name := "target_name=integrity"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ dev_name := "dev_name=" <device_name_str>
+ start := "start=" <N>
+ tag_size := "tag_size=" <N>
+ mode := "mode=" <integrity_mode_str>
+ integrity_mode_str := "J" | "B" | "D" | "R"
+ meta_device := "meta_device=" <meta_device_str>
+ block_size := "block_size=" <N>
+ recalculate := "recalculate=" <yes_no>
+ allow_discards := "allow_discards=" <yes_no>
+ fix_padding := "fix_padding=" <yes_no>
+ fix_hmac := "fix_hmac=" <yes_no>
+ legacy_recalculate := "legacy_recalculate=" <yes_no>
+ journal_sectors := "journal_sectors=" <N>
+ interleave_sectors := "interleave_sectors=" <N>
+ buffer_sectors := "buffer_sectors=" <N>
+ yes_no := "y" | "n"
+
+ E.g.
+ When a 'integrity' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'integrity' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=integrity1,uuid=,major=253,minor=1,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=7856,target_name=integrity,target_version=1.10.0,
+ dev_name=253:0,start=0,tag_size=32,mode=J,recalculate=n,allow_discards=n,fix_padding=n,
+ fix_hmac=n,legacy_recalculate=n,journal_sectors=88,interleave_sectors=32768,buffer_sectors=128;
+
+
+4. linear
+----------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'linear' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <device_name> <,> <start> ";"
+
+ target_name := "target_name=linear"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ device_name := "device_name=" <linear_device_name_str>
+ start := "start=" <N>
+
+ E.g.
+ When a 'linear' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'linear' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=linear1,uuid=linear_uuid1,major=253,minor=2,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=28672,target_name=linear,target_version=1.4.0,
+ device_name=253:1,start=2048;
+
+5. mirror
+----------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'mirror' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <nr_mirrors> ","
+ <mirror_device_data> "," <handle_errors> "," <keep_log> "," <log_type_status> ";"
+
+ target_name := "target_name=mirror"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ nr_mirrors := "nr_mirrors=" <NR>
+ mirror_device_data := <mirror_device_row> | <mirror_device_data><mirror_device_row>
+ mirror_device_row is repeated <NR> times - for <NR> described in <nr_mirrors>.
+ mirror_device_row := <mirror_device_name> "," <mirror_device_status>
+ mirror_device_name := "mirror_device_" <X> "=" <mirror_device_name_str>
+ where <X> ranges from 0 to (<NR> -1) - for <NR> described in <nr_mirrors>.
+ mirror_device_status := "mirror_device_" <X> "_status=" <mirror_device_status_char>
+ where <X> ranges from 0 to (<NR> -1) - for <NR> described in <nr_mirrors>.
+ mirror_device_status_char := "A" | "F" | "D" | "S" | "R" | "U"
+ handle_errors := "handle_errors=" <yes_no>
+ keep_log := "keep_log=" <yes_no>
+ log_type_status := "log_type_status=" <log_type_status_str>
+ yes_no := "y" | "n"
+
+ E.g.
+ When a 'mirror' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'mirror' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=mirror1,uuid=mirror_uuid1,major=253,minor=6,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=2048,target_name=mirror,target_version=1.14.0,nr_mirrors=2,
+ mirror_device_0=253:4,mirror_device_0_status=A,
+ mirror_device_1=253:5,mirror_device_1_status=A,
+ handle_errors=y,keep_log=n,log_type_status=;
+
+6. multipath
+-------------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'multipath' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <nr_priority_groups>
+ ["," <pg_state> "," <priority_groups> "," <priority_group_paths>] ";"
+
+ target_name := "target_name=multipath"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ nr_priority_groups := "nr_priority_groups=" <NPG>
+ priority_groups := <priority_groups_row>|<priority_groups_row><priority_groups>
+ priority_groups_row := "pg_state_" <X> "=" <pg_state_str> "," "nr_pgpaths_" <X> "=" <NPGP> ","
+ "path_selector_name_" <X> "=" <string> "," <priority_group_paths>
+ where <X> ranges from 0 to (<NPG> -1) - for <NPG> described in <nr_priority_groups>.
+ pg_state_str := "E" | "A" | "D"
+ <priority_group_paths> := <priority_group_paths_row> | <priority_group_paths_row><priority_group_paths>
+ priority_group_paths_row := "path_name_" <X> "_" <Y> "=" <string> "," "is_active_" <X> "_" <Y> "=" <is_active_str>
+ "fail_count_" <X> "_" <Y> "=" <N> "," "path_selector_status_" <X> "_" <Y> "=" <path_selector_status_str>
+ where <X> ranges from 0 to (<NPG> -1) - for <NPG> described in <nr_priority_groups>,
+ and <Y> ranges from 0 to (<NPGP> -1) - for <NPGP> described in <priority_groups_row>.
+ is_active_str := "A" | "F"
+
+ E.g.
+ When a 'multipath' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'multipath' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=mp,uuid=,major=253,minor=0,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=2097152,target_name=multipath,target_version=1.14.0,nr_priority_groups=2,
+ pg_state_0=E,nr_pgpaths_0=2,path_selector_name_0=queue-length,
+ path_name_0_0=8:16,is_active_0_0=A,fail_count_0_0=0,path_selector_status_0_0=,
+ path_name_0_1=8:32,is_active_0_1=A,fail_count_0_1=0,path_selector_status_0_1=,
+ pg_state_1=E,nr_pgpaths_1=2,path_selector_name_1=queue-length,
+ path_name_1_0=8:48,is_active_1_0=A,fail_count_1_0=0,path_selector_status_1_0=,
+ path_name_1_1=8:64,is_active_1_1=A,fail_count_1_1=0,path_selector_status_1_1=;
+
+7. raid
+--------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'raid' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <raid_type> "," <raid_disks> "," <raid_state>
+ <raid_device_status> ["," journal_dev_mode] ";"
+
+ target_name := "target_name=raid"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ raid_type := "raid_type=" <raid_type_str>
+ raid_disks := "raid_disks=" <NRD>
+ raid_state := "raid_state=" <raid_state_str>
+ raid_state_str := "frozen" | "reshape" |"resync" | "check" | "repair" | "recover" | "idle" |"undef"
+ raid_device_status := <raid_device_status_row> | <raid_device_status_row><raid_device_status>
+ <raid_device_status_row> is repeated <NRD> times - for <NRD> described in <raid_disks>.
+ raid_device_status_row := "raid_device_" <X> "_status=" <raid_device_status_str>
+ where <X> ranges from 0 to (<NRD> -1) - for <NRD> described in <raid_disks>.
+ raid_device_status_str := "A" | "D" | "a" | "-"
+ journal_dev_mode := "journal_dev_mode=" <journal_dev_mode_str>
+ journal_dev_mode_str := "writethrough" | "writeback" | "invalid"
+
+ E.g.
+ When a 'raid' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'raid' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=raid_LV1,uuid=uuid_raid_LV1,major=253,minor=12,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=2048,target_name=raid,target_version=1.15.1,
+ raid_type=raid10,raid_disks=4,raid_state=idle,
+ raid_device_0_status=A,
+ raid_device_1_status=A,
+ raid_device_2_status=A,
+ raid_device_3_status=A;
+
+
+8. snapshot
+------------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'snapshot' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <snap_origin_name> ","
+ <snap_cow_name> "," <snap_valid> "," <snap_merge_failed> "," <snapshot_overflowed> ";"
+
+ target_name := "target_name=snapshot"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ snap_origin_name := "snap_origin_name=" <string>
+ snap_cow_name := "snap_cow_name=" <string>
+ snap_valid := "snap_valid=" <yes_no>
+ snap_merge_failed := "snap_merge_failed=" <yes_no>
+ snapshot_overflowed := "snapshot_overflowed=" <yes_no>
+ yes_no := "y" | "n"
+
+ E.g.
+ When a 'snapshot' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'snapshot' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=snap1,uuid=snap_uuid1,major=253,minor=13,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=4096,target_name=snapshot,target_version=1.16.0,
+ snap_origin_name=253:11,snap_cow_name=253:12,snap_valid=y,snap_merge_failed=n,snapshot_overflowed=n;
+
+9. striped
+-----------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'striped' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <stripes> "," <chunk_size> ","
+ <stripe_data> ";"
+
+ target_name := "target_name=striped"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ stripes := "stripes=" <NS>
+ chunk_size := "chunk_size=" <N>
+ stripe_data := <stripe_data_row>|<stripe_data><stripe_data_row>
+ stripe_data_row := <stripe_device_name> "," <stripe_physical_start> "," <stripe_status>
+ stripe_device_name := "stripe_" <X> "_device_name=" <stripe_device_name_str>
+ where <X> ranges from 0 to (<NS> -1) - for <NS> described in <stripes>.
+ stripe_physical_start := "stripe_" <X> "_physical_start=" <N>
+ where <X> ranges from 0 to (<NS> -1) - for <NS> described in <stripes>.
+ stripe_status := "stripe_" <X> "_status=" <stripe_status_str>
+ where <X> ranges from 0 to (<NS> -1) - for <NS> described in <stripes>.
+ stripe_status_str := "D" | "A"
+
+ E.g.
+ When a 'striped' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'striped' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=striped1,uuid=striped_uuid1,major=253,minor=5,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=640,target_name=striped,target_version=1.6.0,stripes=2,chunk_size=64,
+ stripe_0_device_name=253:0,stripe_0_physical_start=2048,stripe_0_status=A,
+ stripe_1_device_name=253:3,stripe_1_physical_start=2048,stripe_1_status=A;
+
+10. verity
+----------
+The 'target_attributes' (described as part of EVENT_DATA in 'Table load'
+section above) has the following data format for 'verity' target.
+
+::
+
+ target_attributes := <target_name> "," <target_version> "," <hash_failed> "," <verity_version> ","
+ <data_device_name> "," <hash_device_name> "," <verity_algorithm> "," <root_digest> ","
+ <salt> "," <ignore_zero_blocks> "," <check_at_most_once> ["," <root_hash_sig_key_desc>]
+ ["," <verity_mode>] ";"
+
+ target_name := "target_name=verity"
+ target_version := "target_version=" <N> "." <N> "." <N>
+ hash_failed := "hash_failed=" <hash_failed_str>
+ hash_failed_str := "C" | "V"
+ verity_version := "verity_version=" <verity_version_str>
+ data_device_name := "data_device_name=" <data_device_name_str>
+ hash_device_name := "hash_device_name=" <hash_device_name_str>
+ verity_algorithm := "verity_algorithm=" <verity_algorithm_str>
+ root_digest := "root_digest=" <root_digest_str>
+ salt := "salt=" <salt_str>
+ salt_str := "-" <verity_salt_str>
+ ignore_zero_blocks := "ignore_zero_blocks=" <yes_no>
+ check_at_most_once := "check_at_most_once=" <yes_no>
+ root_hash_sig_key_desc := "root_hash_sig_key_desc="
+ verity_mode := "verity_mode=" <verity_mode_str>
+ verity_mode_str := "ignore_corruption" | "restart_on_corruption" | "panic_on_corruption" | "invalid"
+ yes_no := "y" | "n"
+
+ E.g.
+ When a 'verity' target is loaded, then IMA ASCII measurement log will have an entry
+ similar to the following, depicting what 'verity' attributes are measured in EVENT_DATA
+ for 'dm_table_load' event.
+ (converted from ASCII to text for readability)
+
+ dm_version=4.45.0;
+ name=test-verity,uuid=,major=253,minor=2,minor_count=1,num_targets=1;
+ target_index=0,target_begin=0,target_len=1953120,target_name=verity,target_version=1.8.0,hash_failed=V,
+ verity_version=1,data_device_name=253:1,hash_device_name=253:0,verity_algorithm=sha256,
+ root_digest=29cb87e60ce7b12b443ba6008266f3e41e93e403d7f298f8e3f316b29ff89c5e,
+ salt=e48da609055204e89ae53b655ca2216dd983cf3cb829f34f63a297d106d53e2d,
+ ignore_zero_blocks=n,check_at_most_once=n;
dm-dust
dm-ebs
dm-flakey
+ dm-ima
dm-init
dm-integrity
dm-io
2. the number of blocks
3. the number of free blocks
4. the number of blocks under writeback
+5. the number of read requests
+6. the number of read requests that hit the cache
+7. the number of write requests
+8. the number of write requests that hit uncommitted block
+9. the number of write requests that hit committed block
+10. the number of write requests that bypass the cache
+11. the number of write requests that are allocated in the cache
+12. the number of write requests that are blocked on the freelist
+13. the number of flush requests
+14. the number of discard requests
Messages:
flush
- flush the cache device. The message returns successfully
+ Flush the cache device. The message returns successfully
if the cache device was flushed without an error
flush_on_suspend
- flush the cache device on next suspend. Use this message
+ Flush the cache device on next suspend. Use this message
when you are going to remove the cache device. The proper
sequence for removing the cache device is:
6. the cache device is now inactive and it can be deleted
cleaner
See above "cleaner" constructor documentation.
+ clear_stats
+ Clear the statistics that are reported on the status line
dm-mod-objs += dm-zone.o
endif
+ifeq ($(CONFIG_IMA),y)
+dm-mod-objs += dm-ima.o
+endif
+
ifeq ($(CONFIG_DM_VERITY_FEC),y)
dm-verity-objs += dm-verity-fec.o
endif
DMEMIT(" %s", cache->ctr_args[i]);
if (cache->nr_ctr_args)
DMEMIT(" %s", cache->ctr_args[cache->nr_ctr_args - 1]);
+ break;
+
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ if (get_cache_mode(cache) == CM_FAIL)
+ DMEMIT(",metadata_mode=fail");
+ else if (get_cache_mode(cache) == CM_READ_ONLY)
+ DMEMIT(",metadata_mode=ro");
+ else
+ DMEMIT(",metadata_mode=rw");
+
+ format_dev_t(buf, cache->metadata_dev->bdev->bd_dev);
+ DMEMIT(",cache_metadata_device=%s", buf);
+ format_dev_t(buf, cache->cache_dev->bdev->bd_dev);
+ DMEMIT(",cache_device=%s", buf);
+ format_dev_t(buf, cache->origin_dev->bdev->bd_dev);
+ DMEMIT(",cache_origin_device=%s", buf);
+ DMEMIT(",writethrough=%c", writethrough_mode(cache) ? 'y' : 'n');
+ DMEMIT(",writeback=%c", writeback_mode(cache) ? 'y' : 'n');
+ DMEMIT(",passthrough=%c", passthrough_mode(cache) ? 'y' : 'n');
+ DMEMIT(",metadata2=%c", cache->features.metadata_version == 2 ? 'y' : 'n');
+ DMEMIT(",no_discard_passdown=%c", cache->features.discard_passdown ? 'n' : 'y');
+ DMEMIT(";");
+ break;
}
return;
for (i = 0; i < clone->nr_ctr_args; i++)
DMEMIT(" %s", clone->ctr_args[i]);
+ break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return;
#include <trace/events/block.h>
#include "dm.h"
+#include "dm-ima.h"
#define DM_RESERVED_MAX_IOS 1024
unsigned int nr_zones;
unsigned int *zwp_offset;
#endif
+
+#ifdef CONFIG_IMA
+ struct dm_ima_measurements ima;
+#endif
};
/*
if ((bio_data_dir(io->base_bio) == READ && test_bit(DM_CRYPT_NO_READ_WORKQUEUE, &cc->flags)) ||
(bio_data_dir(io->base_bio) == WRITE && test_bit(DM_CRYPT_NO_WRITE_WORKQUEUE, &cc->flags))) {
/*
- * in_irq(): Crypto API's skcipher_walk_first() refuses to work in hard IRQ context.
+ * in_hardirq(): Crypto API's skcipher_walk_first() refuses to work in hard IRQ context.
* irqs_disabled(): the kernel may run some IO completion from the idle thread, but
* it is being executed with irqs disabled.
*/
- if (in_irq() || irqs_disabled()) {
+ if (in_hardirq() || irqs_disabled()) {
tasklet_init(&io->tasklet, kcryptd_crypt_tasklet, (unsigned long)&io->work);
tasklet_schedule(&io->tasklet);
return;
struct crypt_config *cc = pool_data;
struct page *page;
- if (unlikely(percpu_counter_compare(&cc->n_allocated_pages, dm_crypt_pages_per_client) >= 0) &&
+ /*
+ * Note, percpu_counter_read_positive() may over (and under) estimate
+ * the current usage by at most (batch - 1) * num_online_cpus() pages,
+ * but avoids potential spinlock contention of an exact result.
+ */
+ if (unlikely(percpu_counter_read_positive(&cc->n_allocated_pages) >= dm_crypt_pages_per_client) &&
likely(gfp_mask & __GFP_NORETRY))
return NULL;
if (test_bit(CRYPT_IV_LARGE_SECTORS, &cc->cipher_flags))
DMEMIT(" iv_large_sectors");
}
+ break;
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",allow_discards=%c", ti->num_discard_bios ? 'y' : 'n');
+ DMEMIT(",same_cpu_crypt=%c", test_bit(DM_CRYPT_SAME_CPU, &cc->flags) ? 'y' : 'n');
+ DMEMIT(",submit_from_crypt_cpus=%c", test_bit(DM_CRYPT_NO_OFFLOAD, &cc->flags) ?
+ 'y' : 'n');
+ DMEMIT(",no_read_workqueue=%c", test_bit(DM_CRYPT_NO_READ_WORKQUEUE, &cc->flags) ?
+ 'y' : 'n');
+ DMEMIT(",no_write_workqueue=%c", test_bit(DM_CRYPT_NO_WRITE_WORKQUEUE, &cc->flags) ?
+ 'y' : 'n');
+ DMEMIT(",iv_large_sectors=%c", test_bit(CRYPT_IV_LARGE_SECTORS, &cc->cipher_flags) ?
+ 'y' : 'n');
+
+ if (cc->on_disk_tag_size)
+ DMEMIT(",integrity_tag_size=%u,cipher_auth=%s",
+ cc->on_disk_tag_size, cc->cipher_auth);
+ if (cc->sector_size != (1 << SECTOR_SHIFT))
+ DMEMIT(",sector_size=%d", cc->sector_size);
+ if (cc->cipher_string)
+ DMEMIT(",cipher_string=%s", cc->cipher_string);
+
+ DMEMIT(",key_size=%u", cc->key_size);
+ DMEMIT(",key_parts=%u", cc->key_parts);
+ DMEMIT(",key_extra_size=%u", cc->key_extra_size);
+ DMEMIT(",key_mac_size=%u", cc->key_mac_size);
+ DMEMIT(";");
break;
}
}
DMEMIT_DELAY_CLASS(&dc->flush);
}
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
DMEMIT("%s %llu %u", dd->dev->name,
(unsigned long long)dd->start, dd->blksz);
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
snprintf(result, maxlen, ec->u_bs_set ? "%s %llu %u %u" : "%s %llu %u",
ec->dev->name, (unsigned long long) ec->start, ec->e_bs, ec->u_bs);
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
format_dev_t(buf, era->origin_dev->bdev->bd_dev);
DMEMIT("%s %u", buf, era->sectors_per_block);
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return;
fc->corrupt_bio_value, fc->corrupt_bio_flags);
break;
+
+ case STATUSTYPE_IMA:
+ result[0] = '\0';
+ break;
}
}
--- /dev/null
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2021 Microsoft Corporation
+ *
+ * Author: Tushar Sugandhi <tusharsu@linux.microsoft.com>
+ *
+ * File: dm-ima.c
+ * Enables IMA measurements for DM targets
+ */
+
+#include "dm-core.h"
+#include "dm-ima.h"
+
+#include <linux/ima.h>
+#include <crypto/hash.h>
+#include <linux/crypto.h>
+#include <crypto/hash_info.h>
+
+#define DM_MSG_PREFIX "ima"
+
+/*
+ * Internal function to prefix separator characters in input buffer with escape
+ * character, so that they don't interfere with the construction of key-value pairs,
+ * and clients can split the key1=val1,key2=val2,key3=val3; pairs properly.
+ */
+static void fix_separator_chars(char **buf)
+{
+ int l = strlen(*buf);
+ int i, j, sp = 0;
+
+ for (i = 0; i < l; i++)
+ if ((*buf)[i] == '\\' || (*buf)[i] == ';' || (*buf)[i] == '=' || (*buf)[i] == ',')
+ sp++;
+
+ if (!sp)
+ return;
+
+ for (i = l-1, j = i+sp; i >= 0; i--) {
+ (*buf)[j--] = (*buf)[i];
+ if ((*buf)[i] == '\\' || (*buf)[i] == ';' || (*buf)[i] == '=' || (*buf)[i] == ',')
+ (*buf)[j--] = '\\';
+ }
+}
+
+/*
+ * Internal function to allocate memory for IMA measurements.
+ */
+static void *dm_ima_alloc(size_t len, gfp_t flags, bool noio)
+{
+ unsigned int noio_flag;
+ void *ptr;
+
+ if (noio)
+ noio_flag = memalloc_noio_save();
+
+ ptr = kzalloc(len, flags);
+
+ if (noio)
+ memalloc_noio_restore(noio_flag);
+
+ return ptr;
+}
+
+/*
+ * Internal function to allocate and copy name and uuid for IMA measurements.
+ */
+static int dm_ima_alloc_and_copy_name_uuid(struct mapped_device *md, char **dev_name,
+ char **dev_uuid, bool noio)
+{
+ int r;
+ *dev_name = dm_ima_alloc(DM_NAME_LEN*2, GFP_KERNEL, noio);
+ if (!(*dev_name)) {
+ r = -ENOMEM;
+ goto error;
+ }
+
+ *dev_uuid = dm_ima_alloc(DM_UUID_LEN*2, GFP_KERNEL, noio);
+ if (!(*dev_uuid)) {
+ r = -ENOMEM;
+ goto error;
+ }
+
+ r = dm_copy_name_and_uuid(md, *dev_name, *dev_uuid);
+ if (r)
+ goto error;
+
+ fix_separator_chars(dev_name);
+ fix_separator_chars(dev_uuid);
+
+ return 0;
+error:
+ kfree(*dev_name);
+ kfree(*dev_uuid);
+ *dev_name = NULL;
+ *dev_uuid = NULL;
+ return r;
+}
+
+/*
+ * Internal function to allocate and copy device data for IMA measurements.
+ */
+static int dm_ima_alloc_and_copy_device_data(struct mapped_device *md, char **device_data,
+ unsigned int num_targets, bool noio)
+{
+ char *dev_name = NULL, *dev_uuid = NULL;
+ int r;
+
+ r = dm_ima_alloc_and_copy_name_uuid(md, &dev_name, &dev_uuid, noio);
+ if (r)
+ return r;
+
+ *device_data = dm_ima_alloc(DM_IMA_DEVICE_BUF_LEN, GFP_KERNEL, noio);
+ if (!(*device_data)) {
+ r = -ENOMEM;
+ goto error;
+ }
+
+ scnprintf(*device_data, DM_IMA_DEVICE_BUF_LEN,
+ "name=%s,uuid=%s,major=%d,minor=%d,minor_count=%d,num_targets=%u;",
+ dev_name, dev_uuid, md->disk->major, md->disk->first_minor,
+ md->disk->minors, num_targets);
+error:
+ kfree(dev_name);
+ kfree(dev_uuid);
+ return r;
+}
+
+/*
+ * Internal wrapper function to call IMA to measure DM data.
+ */
+static void dm_ima_measure_data(const char *event_name, const void *buf, size_t buf_len,
+ bool noio)
+{
+ unsigned int noio_flag;
+
+ if (noio)
+ noio_flag = memalloc_noio_save();
+
+ ima_measure_critical_data(DM_NAME, event_name, buf, buf_len, false);
+
+ if (noio)
+ memalloc_noio_restore(noio_flag);
+}
+
+/*
+ * Internal function to allocate and copy current device capacity for IMA measurements.
+ */
+static int dm_ima_alloc_and_copy_capacity_str(struct mapped_device *md, char **capacity_str,
+ bool noio)
+{
+ sector_t capacity;
+
+ capacity = get_capacity(md->disk);
+
+ *capacity_str = dm_ima_alloc(DM_IMA_DEVICE_CAPACITY_BUF_LEN, GFP_KERNEL, noio);
+ if (!(*capacity_str))
+ return -ENOMEM;
+
+ scnprintf(*capacity_str, DM_IMA_DEVICE_BUF_LEN, "current_device_capacity=%llu;",
+ capacity);
+
+ return 0;
+}
+
+/*
+ * Initialize/reset the dm ima related data structure variables.
+ */
+void dm_ima_reset_data(struct mapped_device *md)
+{
+ memset(&(md->ima), 0, sizeof(md->ima));
+ md->ima.dm_version_str_len = strlen(DM_IMA_VERSION_STR);
+}
+
+/*
+ * Build up the IMA data for each target, and finally measure.
+ */
+void dm_ima_measure_on_table_load(struct dm_table *table, unsigned int status_flags)
+{
+ size_t device_data_buf_len, target_metadata_buf_len, target_data_buf_len, l = 0;
+ char *target_metadata_buf = NULL, *target_data_buf = NULL, *digest_buf = NULL;
+ char *ima_buf = NULL, *device_data_buf = NULL;
+ int digest_size, last_target_measured = -1, r;
+ status_type_t type = STATUSTYPE_IMA;
+ size_t cur_total_buf_len = 0;
+ unsigned int num_targets, i;
+ SHASH_DESC_ON_STACK(shash, NULL);
+ struct crypto_shash *tfm = NULL;
+ u8 *digest = NULL;
+ bool noio = false;
+ /*
+ * In below hash_alg_prefix_len assignment +1 is for the additional char (':'),
+ * when prefixing the hash value with the hash algorithm name. e.g. sha256:<hash_value>.
+ */
+ const size_t hash_alg_prefix_len = strlen(DM_IMA_TABLE_HASH_ALG) + 1;
+ char table_load_event_name[] = "dm_table_load";
+
+ ima_buf = dm_ima_alloc(DM_IMA_MEASUREMENT_BUF_LEN, GFP_KERNEL, noio);
+ if (!ima_buf)
+ return;
+
+ target_metadata_buf = dm_ima_alloc(DM_IMA_TARGET_METADATA_BUF_LEN, GFP_KERNEL, noio);
+ if (!target_metadata_buf)
+ goto error;
+
+ target_data_buf = dm_ima_alloc(DM_IMA_TARGET_DATA_BUF_LEN, GFP_KERNEL, noio);
+ if (!target_data_buf)
+ goto error;
+
+ num_targets = dm_table_get_num_targets(table);
+
+ if (dm_ima_alloc_and_copy_device_data(table->md, &device_data_buf, num_targets, noio))
+ goto error;
+
+ tfm = crypto_alloc_shash(DM_IMA_TABLE_HASH_ALG, 0, 0);
+ if (IS_ERR(tfm))
+ goto error;
+
+ shash->tfm = tfm;
+ digest_size = crypto_shash_digestsize(tfm);
+ digest = dm_ima_alloc(digest_size, GFP_KERNEL, noio);
+ if (!digest)
+ goto error;
+
+ r = crypto_shash_init(shash);
+ if (r)
+ goto error;
+
+ memcpy(ima_buf + l, DM_IMA_VERSION_STR, table->md->ima.dm_version_str_len);
+ l += table->md->ima.dm_version_str_len;
+
+ device_data_buf_len = strlen(device_data_buf);
+ memcpy(ima_buf + l, device_data_buf, device_data_buf_len);
+ l += device_data_buf_len;
+
+ for (i = 0; i < num_targets; i++) {
+ struct dm_target *ti = dm_table_get_target(table, i);
+
+ if (!ti)
+ goto error;
+
+ last_target_measured = 0;
+
+ /*
+ * First retrieve the target metadata.
+ */
+ scnprintf(target_metadata_buf, DM_IMA_TARGET_METADATA_BUF_LEN,
+ "target_index=%d,target_begin=%llu,target_len=%llu,",
+ i, ti->begin, ti->len);
+ target_metadata_buf_len = strlen(target_metadata_buf);
+
+ /*
+ * Then retrieve the actual target data.
+ */
+ if (ti->type->status)
+ ti->type->status(ti, type, status_flags, target_data_buf,
+ DM_IMA_TARGET_DATA_BUF_LEN);
+ else
+ target_data_buf[0] = '\0';
+
+ target_data_buf_len = strlen(target_data_buf);
+
+ /*
+ * Check if the total data can fit into the IMA buffer.
+ */
+ cur_total_buf_len = l + target_metadata_buf_len + target_data_buf_len;
+
+ /*
+ * IMA measurements for DM targets are best-effort.
+ * If the total data buffered so far, including the current target,
+ * is too large to fit into DM_IMA_MEASUREMENT_BUF_LEN, measure what
+ * we have in the current buffer, and continue measuring the remaining
+ * targets by prefixing the device metadata again.
+ */
+ if (unlikely(cur_total_buf_len >= DM_IMA_MEASUREMENT_BUF_LEN)) {
+ dm_ima_measure_data(table_load_event_name, ima_buf, l, noio);
+ r = crypto_shash_update(shash, (const u8 *)ima_buf, l);
+ if (r < 0)
+ goto error;
+
+ memset(ima_buf, 0, DM_IMA_MEASUREMENT_BUF_LEN);
+ l = 0;
+
+ /*
+ * Each new "dm_table_load" entry in IMA log should have device data
+ * prefix, so that multiple records from the same "dm_table_load" for
+ * a given device can be linked together.
+ */
+ memcpy(ima_buf + l, DM_IMA_VERSION_STR, table->md->ima.dm_version_str_len);
+ l += table->md->ima.dm_version_str_len;
+
+ memcpy(ima_buf + l, device_data_buf, device_data_buf_len);
+ l += device_data_buf_len;
+
+ /*
+ * If this iteration of the for loop turns out to be the last target
+ * in the table, dm_ima_measure_data("dm_table_load", ...) doesn't need
+ * to be called again, just the hash needs to be finalized.
+ * "last_target_measured" tracks this state.
+ */
+ last_target_measured = 1;
+ }
+
+ /*
+ * Fill-in all the target metadata, so that multiple targets for the same
+ * device can be linked together.
+ */
+ memcpy(ima_buf + l, target_metadata_buf, target_metadata_buf_len);
+ l += target_metadata_buf_len;
+
+ memcpy(ima_buf + l, target_data_buf, target_data_buf_len);
+ l += target_data_buf_len;
+ }
+
+ if (!last_target_measured) {
+ dm_ima_measure_data(table_load_event_name, ima_buf, l, noio);
+
+ r = crypto_shash_update(shash, (const u8 *)ima_buf, l);
+ if (r < 0)
+ goto error;
+ }
+
+ /*
+ * Finalize the table hash, and store it in table->md->ima.inactive_table.hash,
+ * so that the table data can be verified against the future device state change
+ * events, e.g. resume, rename, remove, table-clear etc.
+ */
+ r = crypto_shash_final(shash, digest);
+ if (r < 0)
+ goto error;
+
+ digest_buf = dm_ima_alloc((digest_size*2) + hash_alg_prefix_len + 1, GFP_KERNEL, noio);
+
+ if (!digest_buf)
+ goto error;
+
+ snprintf(digest_buf, hash_alg_prefix_len + 1, "%s:", DM_IMA_TABLE_HASH_ALG);
+
+ for (i = 0; i < digest_size; i++)
+ snprintf((digest_buf + hash_alg_prefix_len + (i*2)), 3, "%02x", digest[i]);
+
+ if (table->md->ima.active_table.hash != table->md->ima.inactive_table.hash)
+ kfree(table->md->ima.inactive_table.hash);
+
+ table->md->ima.inactive_table.hash = digest_buf;
+ table->md->ima.inactive_table.hash_len = strlen(digest_buf);
+ table->md->ima.inactive_table.num_targets = num_targets;
+
+ if (table->md->ima.active_table.device_metadata !=
+ table->md->ima.inactive_table.device_metadata)
+ kfree(table->md->ima.inactive_table.device_metadata);
+
+ table->md->ima.inactive_table.device_metadata = device_data_buf;
+ table->md->ima.inactive_table.device_metadata_len = device_data_buf_len;
+
+ goto exit;
+error:
+ kfree(digest_buf);
+ kfree(device_data_buf);
+exit:
+ kfree(digest);
+ if (tfm)
+ crypto_free_shash(tfm);
+ kfree(ima_buf);
+ kfree(target_metadata_buf);
+ kfree(target_data_buf);
+}
+
+/*
+ * Measure IMA data on device resume.
+ */
+void dm_ima_measure_on_device_resume(struct mapped_device *md, bool swap)
+{
+ char *device_table_data, *dev_name = NULL, *dev_uuid = NULL, *capacity_str = NULL;
+ char active[] = "active_table_hash=";
+ unsigned int active_len = strlen(active), capacity_len = 0;
+ unsigned int l = 0;
+ bool noio = true;
+ bool nodata = true;
+ int r;
+
+ device_table_data = dm_ima_alloc(DM_IMA_DEVICE_BUF_LEN, GFP_KERNEL, noio);
+ if (!device_table_data)
+ return;
+
+ r = dm_ima_alloc_and_copy_capacity_str(md, &capacity_str, noio);
+ if (r)
+ goto error;
+
+ memcpy(device_table_data + l, DM_IMA_VERSION_STR, md->ima.dm_version_str_len);
+ l += md->ima.dm_version_str_len;
+
+ if (swap) {
+ if (md->ima.active_table.hash != md->ima.inactive_table.hash)
+ kfree(md->ima.active_table.hash);
+
+ md->ima.active_table.hash = NULL;
+ md->ima.active_table.hash_len = 0;
+
+ if (md->ima.active_table.device_metadata !=
+ md->ima.inactive_table.device_metadata)
+ kfree(md->ima.active_table.device_metadata);
+
+ md->ima.active_table.device_metadata = NULL;
+ md->ima.active_table.device_metadata_len = 0;
+ md->ima.active_table.num_targets = 0;
+
+ if (md->ima.inactive_table.hash) {
+ md->ima.active_table.hash = md->ima.inactive_table.hash;
+ md->ima.active_table.hash_len = md->ima.inactive_table.hash_len;
+ md->ima.inactive_table.hash = NULL;
+ md->ima.inactive_table.hash_len = 0;
+ }
+
+ if (md->ima.inactive_table.device_metadata) {
+ md->ima.active_table.device_metadata =
+ md->ima.inactive_table.device_metadata;
+ md->ima.active_table.device_metadata_len =
+ md->ima.inactive_table.device_metadata_len;
+ md->ima.active_table.num_targets = md->ima.inactive_table.num_targets;
+ md->ima.inactive_table.device_metadata = NULL;
+ md->ima.inactive_table.device_metadata_len = 0;
+ md->ima.inactive_table.num_targets = 0;
+ }
+ }
+
+ if (md->ima.active_table.device_metadata) {
+ memcpy(device_table_data + l, md->ima.active_table.device_metadata,
+ md->ima.active_table.device_metadata_len);
+ l += md->ima.active_table.device_metadata_len;
+
+ nodata = false;
+ }
+
+ if (md->ima.active_table.hash) {
+ memcpy(device_table_data + l, active, active_len);
+ l += active_len;
+
+ memcpy(device_table_data + l, md->ima.active_table.hash,
+ md->ima.active_table.hash_len);
+ l += md->ima.active_table.hash_len;
+
+ memcpy(device_table_data + l, ";", 1);
+ l++;
+
+ nodata = false;
+ }
+
+ if (nodata) {
+ r = dm_ima_alloc_and_copy_name_uuid(md, &dev_name, &dev_uuid, noio);
+ if (r)
+ goto error;
+
+ scnprintf(device_table_data, DM_IMA_DEVICE_BUF_LEN,
+ "%sname=%s,uuid=%s;device_resume=no_data;",
+ DM_IMA_VERSION_STR, dev_name, dev_uuid);
+ l += strlen(device_table_data);
+
+ }
+
+ capacity_len = strlen(capacity_str);
+ memcpy(device_table_data + l, capacity_str, capacity_len);
+ l += capacity_len;
+
+ dm_ima_measure_data("dm_device_resume", device_table_data, l, noio);
+
+ kfree(dev_name);
+ kfree(dev_uuid);
+error:
+ kfree(capacity_str);
+ kfree(device_table_data);
+}
+
+/*
+ * Measure IMA data on remove.
+ */
+void dm_ima_measure_on_device_remove(struct mapped_device *md, bool remove_all)
+{
+ char *device_table_data, *dev_name = NULL, *dev_uuid = NULL, *capacity_str = NULL;
+ char active_table_str[] = "active_table_hash=";
+ char inactive_table_str[] = "inactive_table_hash=";
+ char device_active_str[] = "device_active_metadata=";
+ char device_inactive_str[] = "device_inactive_metadata=";
+ char remove_all_str[] = "remove_all=";
+ unsigned int active_table_len = strlen(active_table_str);
+ unsigned int inactive_table_len = strlen(inactive_table_str);
+ unsigned int device_active_len = strlen(device_active_str);
+ unsigned int device_inactive_len = strlen(device_inactive_str);
+ unsigned int remove_all_len = strlen(remove_all_str);
+ unsigned int capacity_len = 0;
+ unsigned int l = 0;
+ bool noio = true;
+ bool nodata = true;
+ int r;
+
+ device_table_data = dm_ima_alloc(DM_IMA_DEVICE_BUF_LEN*2, GFP_KERNEL, noio);
+ if (!device_table_data)
+ goto exit;
+
+ r = dm_ima_alloc_and_copy_capacity_str(md, &capacity_str, noio);
+ if (r) {
+ kfree(device_table_data);
+ goto exit;
+ }
+
+ memcpy(device_table_data + l, DM_IMA_VERSION_STR, md->ima.dm_version_str_len);
+ l += md->ima.dm_version_str_len;
+
+ if (md->ima.active_table.device_metadata) {
+ memcpy(device_table_data + l, device_active_str, device_active_len);
+ l += device_active_len;
+
+ memcpy(device_table_data + l, md->ima.active_table.device_metadata,
+ md->ima.active_table.device_metadata_len);
+ l += md->ima.active_table.device_metadata_len;
+
+ nodata = false;
+ }
+
+ if (md->ima.inactive_table.device_metadata) {
+ memcpy(device_table_data + l, device_inactive_str, device_inactive_len);
+ l += device_inactive_len;
+
+ memcpy(device_table_data + l, md->ima.inactive_table.device_metadata,
+ md->ima.inactive_table.device_metadata_len);
+ l += md->ima.inactive_table.device_metadata_len;
+
+ nodata = false;
+ }
+
+ if (md->ima.active_table.hash) {
+ memcpy(device_table_data + l, active_table_str, active_table_len);
+ l += active_table_len;
+
+ memcpy(device_table_data + l, md->ima.active_table.hash,
+ md->ima.active_table.hash_len);
+ l += md->ima.active_table.hash_len;
+
+ memcpy(device_table_data + l, ",", 1);
+ l++;
+
+ nodata = false;
+ }
+
+ if (md->ima.inactive_table.hash) {
+ memcpy(device_table_data + l, inactive_table_str, inactive_table_len);
+ l += inactive_table_len;
+
+ memcpy(device_table_data + l, md->ima.inactive_table.hash,
+ md->ima.inactive_table.hash_len);
+ l += md->ima.inactive_table.hash_len;
+
+ memcpy(device_table_data + l, ",", 1);
+ l++;
+
+ nodata = false;
+ }
+ /*
+ * In case both active and inactive tables, and corresponding
+ * device metadata is cleared/missing - record the name and uuid
+ * in IMA measurements.
+ */
+ if (nodata) {
+ if (dm_ima_alloc_and_copy_name_uuid(md, &dev_name, &dev_uuid, noio))
+ goto error;
+
+ scnprintf(device_table_data, DM_IMA_DEVICE_BUF_LEN,
+ "%sname=%s,uuid=%s;device_remove=no_data;",
+ DM_IMA_VERSION_STR, dev_name, dev_uuid);
+ l += strlen(device_table_data);
+ }
+
+ memcpy(device_table_data + l, remove_all_str, remove_all_len);
+ l += remove_all_len;
+ memcpy(device_table_data + l, remove_all ? "y;" : "n;", 2);
+ l += 2;
+
+ capacity_len = strlen(capacity_str);
+ memcpy(device_table_data + l, capacity_str, capacity_len);
+ l += capacity_len;
+
+ dm_ima_measure_data("dm_device_remove", device_table_data, l, noio);
+
+error:
+ kfree(device_table_data);
+ kfree(capacity_str);
+exit:
+ kfree(md->ima.active_table.device_metadata);
+
+ if (md->ima.active_table.device_metadata !=
+ md->ima.inactive_table.device_metadata)
+ kfree(md->ima.inactive_table.device_metadata);
+
+ kfree(md->ima.active_table.hash);
+
+ if (md->ima.active_table.hash != md->ima.inactive_table.hash)
+ kfree(md->ima.inactive_table.hash);
+
+ dm_ima_reset_data(md);
+
+ kfree(dev_name);
+ kfree(dev_uuid);
+}
+
+/*
+ * Measure ima data on table clear.
+ */
+void dm_ima_measure_on_table_clear(struct mapped_device *md, bool new_map)
+{
+ unsigned int l = 0, capacity_len = 0;
+ char *device_table_data = NULL, *dev_name = NULL, *dev_uuid = NULL, *capacity_str = NULL;
+ char inactive_str[] = "inactive_table_hash=";
+ unsigned int inactive_len = strlen(inactive_str);
+ bool noio = true;
+ bool nodata = true;
+ int r;
+
+ device_table_data = dm_ima_alloc(DM_IMA_DEVICE_BUF_LEN, GFP_KERNEL, noio);
+ if (!device_table_data)
+ return;
+
+ r = dm_ima_alloc_and_copy_capacity_str(md, &capacity_str, noio);
+ if (r)
+ goto error1;
+
+ memcpy(device_table_data + l, DM_IMA_VERSION_STR, md->ima.dm_version_str_len);
+ l += md->ima.dm_version_str_len;
+
+ if (md->ima.inactive_table.device_metadata_len &&
+ md->ima.inactive_table.hash_len) {
+ memcpy(device_table_data + l, md->ima.inactive_table.device_metadata,
+ md->ima.inactive_table.device_metadata_len);
+ l += md->ima.inactive_table.device_metadata_len;
+
+ memcpy(device_table_data + l, inactive_str, inactive_len);
+ l += inactive_len;
+
+ memcpy(device_table_data + l, md->ima.inactive_table.hash,
+ md->ima.inactive_table.hash_len);
+
+ l += md->ima.inactive_table.hash_len;
+
+ memcpy(device_table_data + l, ";", 1);
+ l++;
+
+ nodata = false;
+ }
+
+ if (nodata) {
+ if (dm_ima_alloc_and_copy_name_uuid(md, &dev_name, &dev_uuid, noio))
+ goto error2;
+
+ scnprintf(device_table_data, DM_IMA_DEVICE_BUF_LEN,
+ "%sname=%s,uuid=%s;table_clear=no_data;",
+ DM_IMA_VERSION_STR, dev_name, dev_uuid);
+ l += strlen(device_table_data);
+ }
+
+ capacity_len = strlen(capacity_str);
+ memcpy(device_table_data + l, capacity_str, capacity_len);
+ l += capacity_len;
+
+ dm_ima_measure_data("dm_table_clear", device_table_data, l, noio);
+
+ if (new_map) {
+ if (md->ima.inactive_table.hash &&
+ md->ima.inactive_table.hash != md->ima.active_table.hash)
+ kfree(md->ima.inactive_table.hash);
+
+ md->ima.inactive_table.hash = NULL;
+ md->ima.inactive_table.hash_len = 0;
+
+ if (md->ima.inactive_table.device_metadata &&
+ md->ima.inactive_table.device_metadata != md->ima.active_table.device_metadata)
+ kfree(md->ima.inactive_table.device_metadata);
+
+ md->ima.inactive_table.device_metadata = NULL;
+ md->ima.inactive_table.device_metadata_len = 0;
+ md->ima.inactive_table.num_targets = 0;
+
+ if (md->ima.active_table.hash) {
+ md->ima.inactive_table.hash = md->ima.active_table.hash;
+ md->ima.inactive_table.hash_len = md->ima.active_table.hash_len;
+ }
+
+ if (md->ima.active_table.device_metadata) {
+ md->ima.inactive_table.device_metadata =
+ md->ima.active_table.device_metadata;
+ md->ima.inactive_table.device_metadata_len =
+ md->ima.active_table.device_metadata_len;
+ md->ima.inactive_table.num_targets =
+ md->ima.active_table.num_targets;
+ }
+ }
+
+ kfree(dev_name);
+ kfree(dev_uuid);
+error2:
+ kfree(capacity_str);
+error1:
+ kfree(device_table_data);
+}
+
+/*
+ * Measure IMA data on device rename.
+ */
+void dm_ima_measure_on_device_rename(struct mapped_device *md)
+{
+ char *old_device_data = NULL, *new_device_data = NULL, *combined_device_data = NULL;
+ char *new_dev_name = NULL, *new_dev_uuid = NULL, *capacity_str = NULL;
+ bool noio = true;
+ int r;
+
+ if (dm_ima_alloc_and_copy_device_data(md, &new_device_data,
+ md->ima.active_table.num_targets, noio))
+ return;
+
+ if (dm_ima_alloc_and_copy_name_uuid(md, &new_dev_name, &new_dev_uuid, noio))
+ goto error;
+
+ combined_device_data = dm_ima_alloc(DM_IMA_DEVICE_BUF_LEN * 2, GFP_KERNEL, noio);
+ if (!combined_device_data)
+ goto error;
+
+ r = dm_ima_alloc_and_copy_capacity_str(md, &capacity_str, noio);
+ if (r)
+ goto error;
+
+ old_device_data = md->ima.active_table.device_metadata;
+
+ md->ima.active_table.device_metadata = new_device_data;
+ md->ima.active_table.device_metadata_len = strlen(new_device_data);
+
+ scnprintf(combined_device_data, DM_IMA_DEVICE_BUF_LEN * 2,
+ "%s%snew_name=%s,new_uuid=%s;%s", DM_IMA_VERSION_STR, old_device_data,
+ new_dev_name, new_dev_uuid, capacity_str);
+
+ dm_ima_measure_data("dm_device_rename", combined_device_data, strlen(combined_device_data),
+ noio);
+
+ goto exit;
+
+error:
+ kfree(new_device_data);
+exit:
+ kfree(capacity_str);
+ kfree(combined_device_data);
+ kfree(old_device_data);
+ kfree(new_dev_name);
+ kfree(new_dev_uuid);
+}
--- /dev/null
+/* SPDX-License-Identifier: GPL-2.0
+ *
+ * Copyright (C) 2021 Microsoft Corporation
+ *
+ * Author: Tushar Sugandhi <tusharsu@linux.microsoft.com>
+ *
+ * File: dm-ima.h
+ * Header file for device mapper IMA measurements.
+ */
+
+#ifndef DM_IMA_H
+#define DM_IMA_H
+
+#define DM_IMA_MEASUREMENT_BUF_LEN 4096
+#define DM_IMA_DEVICE_BUF_LEN 1024
+#define DM_IMA_TARGET_METADATA_BUF_LEN 128
+#define DM_IMA_TARGET_DATA_BUF_LEN 2048
+#define DM_IMA_DEVICE_CAPACITY_BUF_LEN 128
+#define DM_IMA_TABLE_HASH_ALG "sha256"
+
+#define __dm_ima_stringify(s) #s
+#define __dm_ima_str(s) __dm_ima_stringify(s)
+
+#define DM_IMA_VERSION_STR "dm_version=" \
+ __dm_ima_str(DM_VERSION_MAJOR) "." \
+ __dm_ima_str(DM_VERSION_MINOR) "." \
+ __dm_ima_str(DM_VERSION_PATCHLEVEL) ";"
+
+#ifdef CONFIG_IMA
+
+struct dm_ima_device_table_metadata {
+ /*
+ * Contains data specific to the device which is common across
+ * all the targets in the table (e.g. name, uuid, major, minor, etc).
+ * The values are stored in comma separated list of key1=val1,key2=val2;
+ * pairs delimited by a semicolon at the end of the list.
+ */
+ char *device_metadata;
+ unsigned int device_metadata_len;
+ unsigned int num_targets;
+
+ /*
+ * Contains the sha256 hashes of the IMA measurements of the target
+ * attributes' key-value pairs from the active/inactive tables.
+ */
+ char *hash;
+ unsigned int hash_len;
+};
+
+/*
+ * This structure contains device metadata, and table hash for
+ * active and inactive tables for ima measurements.
+ */
+struct dm_ima_measurements {
+ struct dm_ima_device_table_metadata active_table;
+ struct dm_ima_device_table_metadata inactive_table;
+ unsigned int dm_version_str_len;
+};
+
+void dm_ima_reset_data(struct mapped_device *md);
+void dm_ima_measure_on_table_load(struct dm_table *table, unsigned int status_flags);
+void dm_ima_measure_on_device_resume(struct mapped_device *md, bool swap);
+void dm_ima_measure_on_device_remove(struct mapped_device *md, bool remove_all);
+void dm_ima_measure_on_table_clear(struct mapped_device *md, bool new_map);
+void dm_ima_measure_on_device_rename(struct mapped_device *md);
+
+#else
+
+static inline void dm_ima_reset_data(struct mapped_device *md) {}
+static inline void dm_ima_measure_on_table_load(struct dm_table *table, unsigned int status_flags) {}
+static inline void dm_ima_measure_on_device_resume(struct mapped_device *md, bool swap) {}
+static inline void dm_ima_measure_on_device_remove(struct mapped_device *md, bool remove_all) {}
+static inline void dm_ima_measure_on_table_clear(struct mapped_device *md, bool new_map) {}
+static inline void dm_ima_measure_on_device_rename(struct mapped_device *md) {}
+
+#endif /* CONFIG_IMA */
+
+#endif /* DM_IMA_H */
EMIT_ALG(journal_mac_alg, "journal_mac");
break;
}
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",dev_name=%s,start=%llu,tag_size=%u,mode=%c",
+ ic->dev->name, ic->start, ic->tag_size, ic->mode);
+
+ if (ic->meta_dev)
+ DMEMIT(",meta_device=%s", ic->meta_dev->name);
+ if (ic->sectors_per_block != 1)
+ DMEMIT(",block_size=%u", ic->sectors_per_block << SECTOR_SHIFT);
+
+ DMEMIT(",recalculate=%c", (ic->sb->flags & cpu_to_le32(SB_FLAG_RECALCULATING)) ?
+ 'y' : 'n');
+ DMEMIT(",allow_discards=%c", ic->discard ? 'y' : 'n');
+ DMEMIT(",fix_padding=%c",
+ ((ic->sb->flags & cpu_to_le32(SB_FLAG_FIXED_PADDING)) != 0) ? 'y' : 'n');
+ DMEMIT(",fix_hmac=%c",
+ ((ic->sb->flags & cpu_to_le32(SB_FLAG_FIXED_HMAC)) != 0) ? 'y' : 'n');
+ DMEMIT(",legacy_recalculate=%c", ic->legacy_recalculate ? 'y' : 'n');
+
+ DMEMIT(",journal_sectors=%u", ic->initial_sectors - SB_SECTORS);
+ DMEMIT(",interleave_sectors=%u", 1U << ic->sb->log2_interleave_sectors);
+ DMEMIT(",buffer_sectors=%u", 1U << ic->log2_buffer_sectors);
+ DMEMIT(";");
+ break;
}
}
*/
#include "dm-core.h"
-
+#include "dm-ima.h"
#include <linux/module.h>
#include <linux/vmalloc.h>
#include <linux/miscdevice.h>
#include <linux/compat.h>
#include <linux/uaccess.h>
+#include <linux/ima.h>
#define DM_MSG_PREFIX "ioctl"
#define DM_DRIVER_EMAIL "dm-devel@redhat.com"
dm_sync_table(md);
dm_table_destroy(t);
}
+ dm_ima_measure_on_device_remove(md, true);
dm_put(md);
if (likely(keep_open_devices))
dm_destroy(md);
param->flags |= DM_UEVENT_GENERATED_FLAG;
md = hc->md;
+
+ dm_ima_measure_on_device_rename(md);
+
up_write(&_hash_lock);
kfree(old_name);
param->flags &= ~DM_DEFERRED_REMOVE;
+ dm_ima_measure_on_device_remove(md, false);
+
if (!dm_kobject_uevent(md, KOBJ_REMOVE, param->event_nr))
param->flags |= DM_UEVENT_GENERATED_FLAG;
if (dm_suspended_md(md)) {
r = dm_resume(md);
- if (!r && !dm_kobject_uevent(md, KOBJ_CHANGE, param->event_nr))
- param->flags |= DM_UEVENT_GENERATED_FLAG;
+ if (!r) {
+ dm_ima_measure_on_device_resume(md, new_map ? true : false);
+
+ if (!dm_kobject_uevent(md, KOBJ_CHANGE, param->event_nr))
+ param->flags |= DM_UEVENT_GENERATED_FLAG;
+ }
}
/*
if (param->flags & DM_STATUS_TABLE_FLAG)
type = STATUSTYPE_TABLE;
+ else if (param->flags & DM_IMA_MEASUREMENT_FLAG)
+ type = STATUSTYPE_IMA;
else
type = STATUSTYPE_INFO;
if (r)
goto err_unlock_md_type;
+ dm_ima_measure_on_table_load(t, STATUSTYPE_IMA);
+
immutable_target_type = dm_get_immutable_target_type(md);
if (immutable_target_type &&
(immutable_target_type != dm_table_get_immutable_target_type(t)) &&
struct hash_cell *hc;
struct mapped_device *md;
struct dm_table *old_map = NULL;
+ bool has_new_map = false;
down_write(&_hash_lock);
if (hc->new_map) {
old_map = hc->new_map;
hc->new_map = NULL;
+ has_new_map = true;
}
param->flags &= ~DM_INACTIVE_PRESENT_FLAG;
dm_sync_table(md);
dm_table_destroy(old_map);
}
+ dm_ima_measure_on_table_clear(md, has_new_map);
dm_put(md);
return 0;
unsigned status_flags, char *result, unsigned maxlen)
{
struct linear_c *lc = (struct linear_c *) ti->private;
+ size_t sz = 0;
switch (type) {
case STATUSTYPE_INFO:
break;
case STATUSTYPE_TABLE:
- snprintf(result, maxlen, "%s %llu", lc->dev->name,
- (unsigned long long)lc->start);
+ DMEMIT("%s %llu", lc->dev->name, (unsigned long long)lc->start);
+ break;
+
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",device_name=%s,start=%llu;", lc->dev->name,
+ (unsigned long long)lc->start);
break;
}
}
DMEMIT("integrated_flush ");
DMEMIT("%s ", table_args);
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return (r) ? 0 : (int)sz;
}
case STATUSTYPE_TABLE:
DMEMIT("%s %s", lc->dev->name, lc->logdev->name);
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
DMEMIT("%s %u %u ", log->type->name,
lc->sync == DEFAULTSYNC ? 1 : 2, lc->region_size);
DMEMIT_SYNC;
+ break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return sz;
lc->sync == DEFAULTSYNC ? 2 : 3, lc->log_dev->name,
lc->region_size);
DMEMIT_SYNC;
+ break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return sz;
static void multipath_status(struct dm_target *ti, status_type_t type,
unsigned status_flags, char *result, unsigned maxlen)
{
- int sz = 0;
+ int sz = 0, pg_counter, pgpath_counter;
unsigned long flags;
struct multipath *m = ti->private;
struct priority_group *pg;
}
}
break;
+
+ case STATUSTYPE_IMA:
+ sz = 0; /*reset the result pointer*/
+
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",nr_priority_groups=%u", m->nr_priority_groups);
+
+ pg_counter = 0;
+ list_for_each_entry(pg, &m->priority_groups, list) {
+ if (pg->bypassed)
+ state = 'D'; /* Disabled */
+ else if (pg == m->current_pg)
+ state = 'A'; /* Currently Active */
+ else
+ state = 'E'; /* Enabled */
+ DMEMIT(",pg_state_%d=%c", pg_counter, state);
+ DMEMIT(",nr_pgpaths_%d=%u", pg_counter, pg->nr_pgpaths);
+ DMEMIT(",path_selector_name_%d=%s", pg_counter, pg->ps.type->name);
+
+ pgpath_counter = 0;
+ list_for_each_entry(p, &pg->pgpaths, list) {
+ DMEMIT(",path_name_%d_%d=%s,is_active_%d_%d=%c,fail_count_%d_%d=%u",
+ pg_counter, pgpath_counter, p->path.dev->name,
+ pg_counter, pgpath_counter, p->is_active ? 'A' : 'F',
+ pg_counter, pgpath_counter, p->fail_count);
+ if (pg->ps.type->status) {
+ DMEMIT(",path_selector_status_%d_%d=",
+ pg_counter, pgpath_counter);
+ sz += pg->ps.type->status(&pg->ps, &p->path,
+ type, result + sz,
+ maxlen - sz);
+ }
+ pgpath_counter++;
+ }
+ pg_counter++;
+ }
+ DMEMIT(";");
+ break;
}
spin_unlock_irqrestore(&m->lock, flags);
case STATUSTYPE_TABLE:
DMEMIT("0 ");
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
pi = path->pscontext;
DMEMIT("%*pb ", cpumask_pr_args(pi->cpumask));
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return sz;
case STATUSTYPE_TABLE:
DMEMIT("%u ", pi->repeat_count);
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
pi = path->pscontext;
DMEMIT("%u ", pi->repeat_count);
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
DMEMIT("%u %u ", pi->repeat_count,
pi->relative_throughput);
break;
+ case STATUSTYPE_IMA:
+ result[0] = '\0';
+ break;
}
}
for (i = 0; i < rs->raid_disks; i++)
DMEMIT(" %s %s", __get_dev_name(rs->dev[i].meta_dev),
__get_dev_name(rs->dev[i].data_dev));
+ break;
+
+ case STATUSTYPE_IMA:
+ rt = get_raid_type_by_ll(mddev->new_level, mddev->new_layout);
+ if (!rt)
+ return;
+
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",raid_type=%s,raid_disks=%d", rt->name, mddev->raid_disks);
+
+ /* Access most recent mddev properties for status output */
+ smp_rmb();
+ recovery = rs->md.recovery;
+ state = decipher_sync_action(mddev, recovery);
+ DMEMIT(",raid_state=%s", sync_str(state));
+
+ for (i = 0; i < rs->raid_disks; i++) {
+ DMEMIT(",raid_device_%d_status=", i);
+ DMEMIT(__raid_dev_status(rs, &rs->dev[i].rdev));
+ }
+
+ if (rt_is_raid456(rt)) {
+ DMEMIT(",journal_dev_mode=");
+ switch (rs->journal_dev.mode) {
+ case R5C_JOURNAL_MODE_WRITE_THROUGH:
+ DMEMIT("%s",
+ _raid456_journal_mode[R5C_JOURNAL_MODE_WRITE_THROUGH].param);
+ break;
+ case R5C_JOURNAL_MODE_WRITE_BACK:
+ DMEMIT("%s",
+ _raid456_journal_mode[R5C_JOURNAL_MODE_WRITE_BACK].param);
+ break;
+ default:
+ DMEMIT("invalid");
+ break;
+ }
+ }
+ DMEMIT(";");
+ break;
}
}
}
break;
+
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",nr_mirrors=%d", ms->nr_mirrors);
+ for (m = 0; m < ms->nr_mirrors; m++) {
+ DMEMIT(",mirror_device_%d=%s", m, ms->mirror[m].dev->name);
+ DMEMIT(",mirror_device_%d_status=%c",
+ m, device_status_char(&(ms->mirror[m])));
+ }
+
+ DMEMIT(",handle_errors=%c", errors_handled(ms) ? 'y' : 'n');
+ DMEMIT(",keep_log=%c", keep_log(ms) ? 'y' : 'n');
+
+ DMEMIT(",log_type_status=");
+ sz += log->type->status(log, type, result+sz, maxlen-sz);
+ DMEMIT(";");
+ break;
}
}
case STATUSTYPE_TABLE:
DMEMIT(" %s %llu", store->userspace_supports_overflow ? "PO" : "P",
(unsigned long long)store->chunk_size);
+ break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return sz;
break;
case STATUSTYPE_TABLE:
DMEMIT(" N %llu", (unsigned long long)store->chunk_size);
+ break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return sz;
DMEMIT(" discard_passdown_origin");
}
break;
+
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",snap_origin_name=%s", snap->origin->name);
+ DMEMIT(",snap_cow_name=%s", snap->cow->name);
+ DMEMIT(",snap_valid=%c", snap->valid ? 'y' : 'n');
+ DMEMIT(",snap_merge_failed=%c", snap->merge_failed ? 'y' : 'n');
+ DMEMIT(",snapshot_overflowed=%c", snap->snapshot_overflowed ? 'y' : 'n');
+ DMEMIT(";");
+ break;
}
}
case STATUSTYPE_TABLE:
snprintf(result, maxlen, "%s", o->dev->name);
break;
+ case STATUSTYPE_IMA:
+ result[0] = '\0';
+ break;
}
}
DMEMIT(" %s %llu", sc->stripe[i].dev->name,
(unsigned long long)sc->stripe[i].physical_start);
break;
+
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",stripes=%d,chunk_size=%llu", sc->stripes,
+ (unsigned long long)sc->chunk_size);
+
+ for (i = 0; i < sc->stripes; i++) {
+ DMEMIT(",stripe_%d_device_name=%s", i, sc->stripe[i].dev->name);
+ DMEMIT(",stripe_%d_physical_start=%llu", i,
+ (unsigned long long)sc->stripe[i].physical_start);
+ DMEMIT(",stripe_%d_status=%c", i,
+ atomic_read(&(sc->stripe[i].error_count)) ? 'D' : 'A');
+ }
+ DMEMIT(";");
+ break;
}
}
DMEMIT(" %s %llu", sctx->path_list[path_nr].dmdev->name,
(unsigned long long)sctx->path_list[path_nr].start);
break;
+
+ case STATUSTYPE_IMA:
+ result[0] = '\0';
+ break;
}
}
(unsigned long long)pt->low_water_blocks);
emit_flags(&pt->requested_pf, result, sz, maxlen);
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return;
if (tc->origin_dev)
DMEMIT(" %s", format_dev_t(buf, tc->origin_dev->bdev->bd_dev));
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
uc->stripes, (unsigned long long)uc->chunk_size, uc->unstripe,
uc->dev->name, (unsigned long long)uc->physical_start);
break;
+
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
DMEMIT(" " DM_VERITY_ROOT_HASH_VERIFICATION_OPT_SIG_KEY
" %s", v->signature_key_desc);
break;
+
+ case STATUSTYPE_IMA:
+ DMEMIT_TARGET_NAME_VERSION(ti->type);
+ DMEMIT(",hash_failed=%c", v->hash_failed ? 'C' : 'V');
+ DMEMIT(",verity_version=%u", v->version);
+ DMEMIT(",data_device_name=%s", v->data_dev->name);
+ DMEMIT(",hash_device_name=%s", v->hash_dev->name);
+ DMEMIT(",verity_algorithm=%s", v->alg_name);
+
+ DMEMIT(",root_digest=");
+ for (x = 0; x < v->digest_size; x++)
+ DMEMIT("%02x", v->root_digest[x]);
+
+ DMEMIT(",salt=");
+ if (!v->salt_size)
+ DMEMIT("-");
+ else
+ for (x = 0; x < v->salt_size; x++)
+ DMEMIT("%02x", v->salt[x]);
+
+ DMEMIT(",ignore_zero_blocks=%c", v->zero_digest ? 'y' : 'n');
+ DMEMIT(",check_at_most_once=%c", v->validated_blocks ? 'y' : 'n');
+ if (v->signature_key_desc)
+ DMEMIT(",root_hash_sig_key_desc=%s", v->signature_key_desc);
+
+ if (v->mode != DM_VERITY_MODE_EIO) {
+ DMEMIT(",verity_mode=");
+ switch (v->mode) {
+ case DM_VERITY_MODE_LOGGING:
+ DMEMIT(DM_VERITY_OPT_LOGGING);
+ break;
+ case DM_VERITY_MODE_RESTART:
+ DMEMIT(DM_VERITY_OPT_RESTART);
+ break;
+ case DM_VERITY_MODE_PANIC:
+ DMEMIT(DM_VERITY_OPT_PANIC);
+ break;
+ default:
+ DMEMIT("invalid");
+ }
+ }
+ DMEMIT(";");
+ break;
}
}
struct bio_set bio_set;
mempool_t copy_pool;
+
+ struct {
+ unsigned long long reads;
+ unsigned long long read_hits;
+ unsigned long long writes;
+ unsigned long long write_hits_uncommitted;
+ unsigned long long write_hits_committed;
+ unsigned long long writes_around;
+ unsigned long long writes_allocate;
+ unsigned long long writes_blocked_on_freelist;
+ unsigned long long flushes;
+ unsigned long long discards;
+ } stats;
};
#define WB_LIST_INLINE 16
return 0;
}
+static int process_clear_stats_mesg(unsigned argc, char **argv, struct dm_writecache *wc)
+{
+ if (argc != 1)
+ return -EINVAL;
+
+ wc_lock(wc);
+ memset(&wc->stats, 0, sizeof wc->stats);
+ wc_unlock(wc);
+
+ return 0;
+}
+
static int writecache_message(struct dm_target *ti, unsigned argc, char **argv,
char *result, unsigned maxlen)
{
r = process_flush_on_suspend_mesg(argc, argv, wc);
else if (!strcasecmp(argv[0], "cleaner"))
r = process_cleaner_mesg(argc, argv, wc);
+ else if (!strcasecmp(argv[0], "clear_stats"))
+ r = process_clear_stats_mesg(argc, argv, wc);
else
DMERR("unrecognised message received: %s", argv[0]);
bio_list_add(&wc->flush_list, bio);
}
-static int writecache_map(struct dm_target *ti, struct bio *bio)
+enum wc_map_op {
+ WC_MAP_SUBMIT,
+ WC_MAP_REMAP,
+ WC_MAP_REMAP_ORIGIN,
+ WC_MAP_RETURN,
+ WC_MAP_ERROR,
+};
+
+static enum wc_map_op writecache_map_remap_origin(struct dm_writecache *wc, struct bio *bio,
+ struct wc_entry *e)
{
- struct wc_entry *e;
- struct dm_writecache *wc = ti->private;
+ if (e) {
+ sector_t next_boundary =
+ read_original_sector(wc, e) - bio->bi_iter.bi_sector;
+ if (next_boundary < bio->bi_iter.bi_size >> SECTOR_SHIFT)
+ dm_accept_partial_bio(bio, next_boundary);
+ }
- bio->bi_private = NULL;
+ return WC_MAP_REMAP_ORIGIN;
+}
- wc_lock(wc);
+static enum wc_map_op writecache_map_read(struct dm_writecache *wc, struct bio *bio)
+{
+ enum wc_map_op map_op;
+ struct wc_entry *e;
- if (unlikely(bio->bi_opf & REQ_PREFLUSH)) {
- if (writecache_has_error(wc))
- goto unlock_error;
+read_next_block:
+ wc->stats.reads++;
+ e = writecache_find_entry(wc, bio->bi_iter.bi_sector, WFE_RETURN_FOLLOWING);
+ if (e && read_original_sector(wc, e) == bio->bi_iter.bi_sector) {
+ wc->stats.read_hits++;
if (WC_MODE_PMEM(wc)) {
- writecache_flush(wc);
- if (writecache_has_error(wc))
- goto unlock_error;
- if (unlikely(wc->cleaner) || unlikely(wc->metadata_only))
- goto unlock_remap_origin;
- goto unlock_submit;
+ bio_copy_block(wc, bio, memory_data(wc, e));
+ if (bio->bi_iter.bi_size)
+ goto read_next_block;
+ map_op = WC_MAP_SUBMIT;
} else {
- if (dm_bio_get_target_bio_nr(bio))
- goto unlock_remap_origin;
- writecache_offload_bio(wc, bio);
- goto unlock_return;
+ dm_accept_partial_bio(bio, wc->block_size >> SECTOR_SHIFT);
+ bio_set_dev(bio, wc->ssd_dev->bdev);
+ bio->bi_iter.bi_sector = cache_sector(wc, e);
+ if (!writecache_entry_is_committed(wc, e))
+ writecache_wait_for_ios(wc, WRITE);
+ map_op = WC_MAP_REMAP;
}
+ } else {
+ map_op = writecache_map_remap_origin(wc, bio, e);
}
- bio->bi_iter.bi_sector = dm_target_offset(ti, bio->bi_iter.bi_sector);
+ return map_op;
+}
- if (unlikely((((unsigned)bio->bi_iter.bi_sector | bio_sectors(bio)) &
- (wc->block_size / 512 - 1)) != 0)) {
- DMERR("I/O is not aligned, sector %llu, size %u, block size %u",
- (unsigned long long)bio->bi_iter.bi_sector,
- bio->bi_iter.bi_size, wc->block_size);
- goto unlock_error;
- }
+static enum wc_map_op writecache_bio_copy_ssd(struct dm_writecache *wc, struct bio *bio,
+ struct wc_entry *e, bool search_used)
+{
+ unsigned bio_size = wc->block_size;
+ sector_t start_cache_sec = cache_sector(wc, e);
+ sector_t current_cache_sec = start_cache_sec + (bio_size >> SECTOR_SHIFT);
- if (unlikely(bio_op(bio) == REQ_OP_DISCARD)) {
- if (writecache_has_error(wc))
- goto unlock_error;
- if (WC_MODE_PMEM(wc)) {
- writecache_discard(wc, bio->bi_iter.bi_sector, bio_end_sector(bio));
- goto unlock_remap_origin;
+ while (bio_size < bio->bi_iter.bi_size) {
+ if (!search_used) {
+ struct wc_entry *f = writecache_pop_from_freelist(wc, current_cache_sec);
+ if (!f)
+ break;
+ write_original_sector_seq_count(wc, f, bio->bi_iter.bi_sector +
+ (bio_size >> SECTOR_SHIFT), wc->seq_count);
+ writecache_insert_entry(wc, f);
+ wc->uncommitted_blocks++;
} else {
- writecache_offload_bio(wc, bio);
- goto unlock_return;
+ struct wc_entry *f;
+ struct rb_node *next = rb_next(&e->rb_node);
+ if (!next)
+ break;
+ f = container_of(next, struct wc_entry, rb_node);
+ if (f != e + 1)
+ break;
+ if (read_original_sector(wc, f) !=
+ read_original_sector(wc, e) + (wc->block_size >> SECTOR_SHIFT))
+ break;
+ if (unlikely(f->write_in_progress))
+ break;
+ if (writecache_entry_is_committed(wc, f))
+ wc->overwrote_committed = true;
+ e = f;
}
+ bio_size += wc->block_size;
+ current_cache_sec += wc->block_size >> SECTOR_SHIFT;
}
- if (bio_data_dir(bio) == READ) {
-read_next_block:
- e = writecache_find_entry(wc, bio->bi_iter.bi_sector, WFE_RETURN_FOLLOWING);
- if (e && read_original_sector(wc, e) == bio->bi_iter.bi_sector) {
- if (WC_MODE_PMEM(wc)) {
- bio_copy_block(wc, bio, memory_data(wc, e));
- if (bio->bi_iter.bi_size)
- goto read_next_block;
- goto unlock_submit;
- } else {
- dm_accept_partial_bio(bio, wc->block_size >> SECTOR_SHIFT);
- bio_set_dev(bio, wc->ssd_dev->bdev);
- bio->bi_iter.bi_sector = cache_sector(wc, e);
- if (!writecache_entry_is_committed(wc, e))
- writecache_wait_for_ios(wc, WRITE);
- goto unlock_remap;
+ bio_set_dev(bio, wc->ssd_dev->bdev);
+ bio->bi_iter.bi_sector = start_cache_sec;
+ dm_accept_partial_bio(bio, bio_size >> SECTOR_SHIFT);
+
+ if (unlikely(wc->uncommitted_blocks >= wc->autocommit_blocks)) {
+ wc->uncommitted_blocks = 0;
+ queue_work(wc->writeback_wq, &wc->flush_work);
+ } else {
+ writecache_schedule_autocommit(wc);
+ }
+
+ return WC_MAP_REMAP;
+}
+
+static enum wc_map_op writecache_map_write(struct dm_writecache *wc, struct bio *bio)
+{
+ struct wc_entry *e;
+
+ do {
+ bool found_entry = false;
+ bool search_used = false;
+ wc->stats.writes++;
+ if (writecache_has_error(wc))
+ return WC_MAP_ERROR;
+ e = writecache_find_entry(wc, bio->bi_iter.bi_sector, 0);
+ if (e) {
+ if (!writecache_entry_is_committed(wc, e)) {
+ wc->stats.write_hits_uncommitted++;
+ search_used = true;
+ goto bio_copy;
}
- } else {
- if (e) {
- sector_t next_boundary =
- read_original_sector(wc, e) - bio->bi_iter.bi_sector;
- if (next_boundary < bio->bi_iter.bi_size >> SECTOR_SHIFT) {
- dm_accept_partial_bio(bio, next_boundary);
- }
+ wc->stats.write_hits_committed++;
+ if (!WC_MODE_PMEM(wc) && !e->write_in_progress) {
+ wc->overwrote_committed = true;
+ search_used = true;
+ goto bio_copy;
}
- goto unlock_remap_origin;
+ found_entry = true;
+ } else {
+ if (unlikely(wc->cleaner) ||
+ (wc->metadata_only && !(bio->bi_opf & REQ_META)))
+ goto direct_write;
}
- } else {
- do {
- bool found_entry = false;
- bool search_used = false;
- if (writecache_has_error(wc))
- goto unlock_error;
- e = writecache_find_entry(wc, bio->bi_iter.bi_sector, 0);
- if (e) {
- if (!writecache_entry_is_committed(wc, e)) {
- search_used = true;
- goto bio_copy;
- }
- if (!WC_MODE_PMEM(wc) && !e->write_in_progress) {
- wc->overwrote_committed = true;
- search_used = true;
- goto bio_copy;
- }
- found_entry = true;
- } else {
- if (unlikely(wc->cleaner) ||
- (wc->metadata_only && !(bio->bi_opf & REQ_META)))
- goto direct_write;
- }
- e = writecache_pop_from_freelist(wc, (sector_t)-1);
- if (unlikely(!e)) {
- if (!WC_MODE_PMEM(wc) && !found_entry) {
+ e = writecache_pop_from_freelist(wc, (sector_t)-1);
+ if (unlikely(!e)) {
+ if (!WC_MODE_PMEM(wc) && !found_entry) {
direct_write:
- e = writecache_find_entry(wc, bio->bi_iter.bi_sector, WFE_RETURN_FOLLOWING);
- if (e) {
- sector_t next_boundary = read_original_sector(wc, e) - bio->bi_iter.bi_sector;
- BUG_ON(!next_boundary);
- if (next_boundary < bio->bi_iter.bi_size >> SECTOR_SHIFT) {
- dm_accept_partial_bio(bio, next_boundary);
- }
- }
- goto unlock_remap_origin;
- }
- writecache_wait_on_freelist(wc);
- continue;
+ wc->stats.writes_around++;
+ e = writecache_find_entry(wc, bio->bi_iter.bi_sector, WFE_RETURN_FOLLOWING);
+ return writecache_map_remap_origin(wc, bio, e);
}
- write_original_sector_seq_count(wc, e, bio->bi_iter.bi_sector, wc->seq_count);
- writecache_insert_entry(wc, e);
- wc->uncommitted_blocks++;
+ wc->stats.writes_blocked_on_freelist++;
+ writecache_wait_on_freelist(wc);
+ continue;
+ }
+ write_original_sector_seq_count(wc, e, bio->bi_iter.bi_sector, wc->seq_count);
+ writecache_insert_entry(wc, e);
+ wc->uncommitted_blocks++;
+ wc->stats.writes_allocate++;
bio_copy:
- if (WC_MODE_PMEM(wc)) {
- bio_copy_block(wc, bio, memory_data(wc, e));
- } else {
- unsigned bio_size = wc->block_size;
- sector_t start_cache_sec = cache_sector(wc, e);
- sector_t current_cache_sec = start_cache_sec + (bio_size >> SECTOR_SHIFT);
-
- while (bio_size < bio->bi_iter.bi_size) {
- if (!search_used) {
- struct wc_entry *f = writecache_pop_from_freelist(wc, current_cache_sec);
- if (!f)
- break;
- write_original_sector_seq_count(wc, f, bio->bi_iter.bi_sector +
- (bio_size >> SECTOR_SHIFT), wc->seq_count);
- writecache_insert_entry(wc, f);
- wc->uncommitted_blocks++;
- } else {
- struct wc_entry *f;
- struct rb_node *next = rb_next(&e->rb_node);
- if (!next)
- break;
- f = container_of(next, struct wc_entry, rb_node);
- if (f != e + 1)
- break;
- if (read_original_sector(wc, f) !=
- read_original_sector(wc, e) + (wc->block_size >> SECTOR_SHIFT))
- break;
- if (unlikely(f->write_in_progress))
- break;
- if (writecache_entry_is_committed(wc, f))
- wc->overwrote_committed = true;
- e = f;
- }
- bio_size += wc->block_size;
- current_cache_sec += wc->block_size >> SECTOR_SHIFT;
- }
+ if (WC_MODE_PMEM(wc))
+ bio_copy_block(wc, bio, memory_data(wc, e));
+ else
+ return writecache_bio_copy_ssd(wc, bio, e, search_used);
+ } while (bio->bi_iter.bi_size);
- bio_set_dev(bio, wc->ssd_dev->bdev);
- bio->bi_iter.bi_sector = start_cache_sec;
- dm_accept_partial_bio(bio, bio_size >> SECTOR_SHIFT);
+ if (unlikely(bio->bi_opf & REQ_FUA || wc->uncommitted_blocks >= wc->autocommit_blocks))
+ writecache_flush(wc);
+ else
+ writecache_schedule_autocommit(wc);
- if (unlikely(wc->uncommitted_blocks >= wc->autocommit_blocks)) {
- wc->uncommitted_blocks = 0;
- queue_work(wc->writeback_wq, &wc->flush_work);
- } else {
- writecache_schedule_autocommit(wc);
- }
- goto unlock_remap;
- }
- } while (bio->bi_iter.bi_size);
+ return WC_MAP_SUBMIT;
+}
- if (unlikely(bio->bi_opf & REQ_FUA ||
- wc->uncommitted_blocks >= wc->autocommit_blocks))
- writecache_flush(wc);
- else
- writecache_schedule_autocommit(wc);
- goto unlock_submit;
+static enum wc_map_op writecache_map_flush(struct dm_writecache *wc, struct bio *bio)
+{
+ if (writecache_has_error(wc))
+ return WC_MAP_ERROR;
+
+ if (WC_MODE_PMEM(wc)) {
+ wc->stats.flushes++;
+ writecache_flush(wc);
+ if (writecache_has_error(wc))
+ return WC_MAP_ERROR;
+ else if (unlikely(wc->cleaner) || unlikely(wc->metadata_only))
+ return WC_MAP_REMAP_ORIGIN;
+ return WC_MAP_SUBMIT;
}
+ /* SSD: */
+ if (dm_bio_get_target_bio_nr(bio))
+ return WC_MAP_REMAP_ORIGIN;
+ wc->stats.flushes++;
+ writecache_offload_bio(wc, bio);
+ return WC_MAP_RETURN;
+}
-unlock_remap_origin:
- if (likely(wc->pause != 0)) {
- if (bio_op(bio) == REQ_OP_WRITE) {
- dm_iot_io_begin(&wc->iot, 1);
- bio->bi_private = (void *)2;
- }
+static enum wc_map_op writecache_map_discard(struct dm_writecache *wc, struct bio *bio)
+{
+ wc->stats.discards++;
+
+ if (writecache_has_error(wc))
+ return WC_MAP_ERROR;
+
+ if (WC_MODE_PMEM(wc)) {
+ writecache_discard(wc, bio->bi_iter.bi_sector, bio_end_sector(bio));
+ return WC_MAP_REMAP_ORIGIN;
}
- bio_set_dev(bio, wc->dev->bdev);
- wc_unlock(wc);
- return DM_MAPIO_REMAPPED;
+ /* SSD: */
+ writecache_offload_bio(wc, bio);
+ return WC_MAP_RETURN;
+}
-unlock_remap:
- /* make sure that writecache_end_io decrements bio_in_progress: */
- bio->bi_private = (void *)1;
- atomic_inc(&wc->bio_in_progress[bio_data_dir(bio)]);
- wc_unlock(wc);
- return DM_MAPIO_REMAPPED;
+static int writecache_map(struct dm_target *ti, struct bio *bio)
+{
+ struct dm_writecache *wc = ti->private;
+ enum wc_map_op map_op;
-unlock_submit:
- wc_unlock(wc);
- bio_endio(bio);
- return DM_MAPIO_SUBMITTED;
+ bio->bi_private = NULL;
-unlock_return:
- wc_unlock(wc);
- return DM_MAPIO_SUBMITTED;
+ wc_lock(wc);
-unlock_error:
- wc_unlock(wc);
- bio_io_error(bio);
- return DM_MAPIO_SUBMITTED;
+ if (unlikely(bio->bi_opf & REQ_PREFLUSH)) {
+ map_op = writecache_map_flush(wc, bio);
+ goto done;
+ }
+
+ bio->bi_iter.bi_sector = dm_target_offset(ti, bio->bi_iter.bi_sector);
+
+ if (unlikely((((unsigned)bio->bi_iter.bi_sector | bio_sectors(bio)) &
+ (wc->block_size / 512 - 1)) != 0)) {
+ DMERR("I/O is not aligned, sector %llu, size %u, block size %u",
+ (unsigned long long)bio->bi_iter.bi_sector,
+ bio->bi_iter.bi_size, wc->block_size);
+ map_op = WC_MAP_ERROR;
+ goto done;
+ }
+
+ if (unlikely(bio_op(bio) == REQ_OP_DISCARD)) {
+ map_op = writecache_map_discard(wc, bio);
+ goto done;
+ }
+
+ if (bio_data_dir(bio) == READ)
+ map_op = writecache_map_read(wc, bio);
+ else
+ map_op = writecache_map_write(wc, bio);
+done:
+ switch (map_op) {
+ case WC_MAP_REMAP_ORIGIN:
+ if (likely(wc->pause != 0)) {
+ if (bio_op(bio) == REQ_OP_WRITE) {
+ dm_iot_io_begin(&wc->iot, 1);
+ bio->bi_private = (void *)2;
+ }
+ }
+ bio_set_dev(bio, wc->dev->bdev);
+ wc_unlock(wc);
+ return DM_MAPIO_REMAPPED;
+
+ case WC_MAP_REMAP:
+ /* make sure that writecache_end_io decrements bio_in_progress: */
+ bio->bi_private = (void *)1;
+ atomic_inc(&wc->bio_in_progress[bio_data_dir(bio)]);
+ wc_unlock(wc);
+ return DM_MAPIO_REMAPPED;
+
+ case WC_MAP_SUBMIT:
+ wc_unlock(wc);
+ bio_endio(bio);
+ return DM_MAPIO_SUBMITTED;
+
+ case WC_MAP_RETURN:
+ wc_unlock(wc);
+ return DM_MAPIO_SUBMITTED;
+
+ case WC_MAP_ERROR:
+ wc_unlock(wc);
+ bio_io_error(bio);
+ return DM_MAPIO_SUBMITTED;
+
+ default:
+ BUG();
+ return -1;
+ }
}
static int writecache_end_io(struct dm_target *ti, struct bio *bio, blk_status_t *status)
switch (type) {
case STATUSTYPE_INFO:
- DMEMIT("%ld %llu %llu %llu", writecache_has_error(wc),
+ DMEMIT("%ld %llu %llu %llu %llu %llu %llu %llu %llu %llu %llu %llu %llu %llu",
+ writecache_has_error(wc),
(unsigned long long)wc->n_blocks, (unsigned long long)wc->freelist_size,
- (unsigned long long)wc->writeback_size);
+ (unsigned long long)wc->writeback_size,
+ wc->stats.reads,
+ wc->stats.read_hits,
+ wc->stats.writes,
+ wc->stats.write_hits_uncommitted,
+ wc->stats.write_hits_committed,
+ wc->stats.writes_around,
+ wc->stats.writes_allocate,
+ wc->stats.writes_blocked_on_freelist,
+ wc->stats.flushes,
+ wc->stats.discards);
break;
case STATUSTYPE_TABLE:
DMEMIT("%c %s %s %u ", WC_MODE_PMEM(wc) ? 'p' : 's',
if (wc->pause_set)
DMEMIT(" pause_writeback %u", wc->pause_value);
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
}
static struct target_type writecache_target = {
.name = "writecache",
- .version = {1, 5, 0},
+ .version = {1, 6, 0},
.module = THIS_MODULE,
.ctr = writecache_ctr,
.dtr = writecache_dtr,
DMEMIT(" %s", buf);
}
break;
+ case STATUSTYPE_IMA:
+ *result = '\0';
+ break;
}
return;
}
#include "dm-core.h"
#include "dm-rq.h"
#include "dm-uevent.h"
+#include "dm-ima.h"
#include <linux/init.h>
#include <linux/module.h>
static int __init dm_init(void)
{
const int count = ARRAY_SIZE(_inits);
-
int r, i;
+#if (IS_ENABLED(CONFIG_IMA) && !IS_ENABLED(CONFIG_IMA_DISABLE_HTABLE))
+ DMWARN("CONFIG_IMA_DISABLE_HTABLE is disabled."
+ " Duplicate IMA measurements will not be recorded in the IMA log.");
+#endif
+
for (i = 0; i < count; i++) {
r = _inits[i]();
if (r)
}
return 0;
-
- bad:
+bad:
while (i--)
_exits[i]();
if (!md)
return -ENXIO;
+ dm_ima_reset_data(md);
+
*result = md;
return 0;
}
DM_TYPE_DAX_BIO_BASED = 3,
};
-typedef enum { STATUSTYPE_INFO, STATUSTYPE_TABLE } status_type_t;
+typedef enum { STATUSTYPE_INFO, STATUSTYPE_TABLE, STATUSTYPE_IMA } status_type_t;
union map_info {
void *ptr;
#define DMEMIT(x...) sz += ((sz >= maxlen) ? \
0 : scnprintf(result + sz, maxlen - sz, x))
+#define DMEMIT_TARGET_NAME_VERSION(y) \
+ DMEMIT("target_name=%s,target_version=%u.%u.%u", \
+ (y)->name, (y)->version[0], (y)->version[1], (y)->version[2])
+
/*
* Definitions of return values from target end_io function.
*/
*/
#define DM_INTERNAL_SUSPEND_FLAG (1 << 18) /* Out */
+/*
+ * If set, returns in the in buffer passed by UM, the raw table information
+ * that would be measured by IMA subsystem on device state change.
+ */
+#define DM_IMA_MEASUREMENT_FLAG (1 << 19) /* In */
+
#endif /* _LINUX_DM_IOCTL_H */
CRITICAL_DATA, 0, event_label,
hash);
}
+EXPORT_SYMBOL_GPL(ima_measure_critical_data);
static int __init init_ima(void)
{
projects / linux-2.6-microblaze.git / commitdiff