aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Fix zfs_arc_max minimum tuningberen122018-04-121-1/+1
| | | | | | | | | | When setting `zfs_arc_max` its minimum value is allowed to be 64 MiB. There was an off-by-1 error which can matter on tiny systems. Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Chris Zubrzycki <[email protected]> Closes #7417
* OpenZFS 9286 - want refreservation=autoMike Gerdts2018-04-119-11/+424
| | | | | | | | | | | | | | | | | | | | Authored by: Mike Gerdts <[email protected]> Reviewed by: Allan Jude <[email protected]> Reviewed by: Matthew Ahrens <[email protected]> Reviewed by: John Kennedy <[email protected]> Reviewed by: Andy Stormont <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Approved by: Richard Lowe <[email protected]> Ported-by: Don Brady <[email protected]> Porting Notes: * Adopted destroy_dataset in ZTS test cleanup * Use ksh shebang instead of bash for new tests OpenZFS-issue: https://www.illumos.org/issues/9286 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/723d0c85 Closes #7387
* Fix zpool set feature@<feature>=disabledLOLi2018-04-115-9/+95
| | | | | | | | | Commit e4010f2 accidentally allows zpool to set pool features to "disabled"; this should only be allowed at pool creation. This commit adds additional checks and test coverage to 'zpool set'. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: loli10K <[email protected]> Closes #7402
* zfs(8): fix dedup omission during mdoc overhaulDeHackEd2018-04-101-0/+30
| | | | | | | | | | | The property description has been updated with new algorithms as well. Reviewed-by: Matt Ahrens <[email protected]> Reviewed-by: Richard Elling <[email protected]> Reviewed-by: loli10K <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: George Melikov <[email protected]> Signed-off-by: DHE <[email protected]> Closes #7377
* Fix race in dnode_check_slots_free()Tom Caputi2018-04-106-11/+43
| | | | | | | | | | | | | | | | | | Currently, dnode_check_slots_free() works by checking dn->dn_type in the dnode to determine if the dnode is reclaimable. However, there is a small window of time between dnode_free_sync() in the first call to dsl_dataset_sync() and when the useraccounting code is run when the type is set DMU_OT_NONE, but the dnode is not yet evictable, leading to crashes. This patch adds the ability for dnodes to track which txg they were last dirtied in and adds a check for this before performing the reclaim. This patch also corrects several instances when dn_dirty_link was treated as a list_node_t when it is technically a multilist_node_t. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7147 Closes #7388
* Linux compat 4.16: blk_queue_flag_{set,clear}Giuseppe Di Natale2018-04-104-4/+60
| | | | | | | | | | queue_flag_{set,clear}_unlocked are now private interfaces in the Linux kernel (https://github.com/torvalds/linux/commit/8a0ac14). Use blk_queue_flag_{set,clear} interfaces which were introduced as of https://github.com/torvalds/linux/commit/8814ce8. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Giuseppe Di Natale <[email protected]> Closes #7410
* Correct swapped keylocation error messagesTom Caputi2018-04-091-4/+4
| | | | | | | | | | | This patch corrects a small issue where two error messages in the code that checks for invalid keylocations were swapped. Reviewed by: Matt Ahrens <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7418
* Revert "Handle zap_add() failures in mixed ... "Tony Hutter2018-04-099-290/+32
| | | | | | | | | | | | This reverts commit cc63068e95ee725cce03b1b7ce50179825a6cda5. Under certain circumstances this change can result in an ENOSPC error when adding new files to a directory. See #7401 for full details. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tony Hutter <[email protected]> Issue #7401 Cloes #7416
* Fix 'zfs send/recv' hang with 16M blocksBrian Behlendorf2018-04-082-4/+37
| | | | | | | | | | | When using 16MB blocks the send/recv queue's aren't quite big enough. This change leaves the default 16M queue size which a good value for most pools. But it additionally ensures that the queue sizes are at least twice the allowed zfs_max_recordsize. Reviewed-by: loli10K <[email protected]> Signed-off-by: Brian Behlendorf <[email protected]> Closes #7365 Closes #7404
* Clean up (k)shlib and cfg file shebangsGiuseppe Di Natale2018-04-0825-30/+27
| | | | | | | | | | | | Most kshlib files are imported by other scripts and do not have a shebang at the top of their files. Make all kshlib follow this convention. Remove shebangs from cfg files as well. Reviewed-by: loli10K <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Giuseppe Di Natale <[email protected]> Close #7406
* Fix "file is executable, but no shebang" warningsTony Hutter2018-04-0683-148/+296
| | | | | | | | | | | | | Fedora 28's RPM build checks warn when executable files don't have a shebang line. These warnings are caused when we (incorrectly) include data & config files in the_SCRIPTS automake lines. Files in _SCRIPTS are marked executable by automake. This patch fixes the issue by including non-executable scripts in a _DATA line instead. Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: Tony Hutter <[email protected]> Closes #7359 Closes #7395
* Exclude python scripts from RPM shebang checkTony Hutter2018-04-061-0/+10
| | | | | | | | | | | | | | | | | | The newest Fedora packaging rules print warnings for scripts using the /usr/bin/python shebang: *** WARNING: mangling shebang in /usr/bin/arc_summary.py from #!/usr/bin/python to #!/usr/bin/python2. This will become an ERROR, fix it manually! Fedora wants all cross compatible scripts to pick python3. Since we don't want our users to have to pick a specific version of python, we exclude our scripts from the RPM build check. Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: Tony Hutter <[email protected]> Closes #7360 Closes #7399
* systemd mount generator and tracking ZEDLETAntonio Russo2018-04-0612-3/+282
| | | | | | | | | | | | | | | | | | | | | | | | zfs-mount-generator implements the "systemd generator" protocol, producing systemd.mount units from the cached outputs of zfs list, during early boot, integrating with systemd. Each pool has an indpendent cache of the command zfs list -H -oname,mountpoint,canmount -tfilesystem -r $pool which is kept synchronized by the ZEDLET history_event-zfs-list-cacher.sh Datasets not in the cache will be loaded later in the boot process by zfs-mount.service, including pools without a cache. Among other things, this allows for complex mount hierarchies. Reviewed-by: Fabian Grünbichler <[email protected]> Reviewed-by: Richard Laager <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Antonio Russo <[email protected]> Closes #7329
* Fixes for SNPRINTF_BLKPTR with encrypted BP'sMatthew Ahrens2018-04-063-17/+27
| | | | | | | | | | | | | mdb doesn't have dmu_ot[], so we need a different mechanism for its SNPRINTF_BLKPTR() to determine if the BP is encrypted vs authenticated. Additionally, since it already relies on BP_IS_ENCRYPTED (etc), SNPRINTF_BLKPTR might as well figure out the "crypt_type" on its own, rather than making the caller do so. Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Tom Caputi <[email protected]> Signed-off-by: Matthew Ahrens <[email protected]> Closes #7390
* Fix divide-by-zero in mmp_delay_update()Olaf Faaland2018-04-061-1/+1
| | | | | | | | | | | | vdev_count_leaves() in the denominator may return 0, caught by Coverity. Introduced by * 533ea04 Update mmp_delay on sync or skipped, failed write Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: George Melikov <[email protected]> Signed-off-by: Olaf Faaland <[email protected]> Closes #7391
* Make encrypted "zfs mount -a" failures consistentTom Caputi2018-04-062-3/+26
| | | | | | | | | | | Currently, "zfs mount -a" will print a warning and fail to mount any encrypted datasets that do not have a key loaded. This patch makes the behavior of this failure consistent with other failure modes ("zfs mount -a" will silently continue, explict "zfs mount" will print a message and return an error code. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7382
* Update mmp_delay on sync or skipped, failed writeOlaf Faaland2018-04-046-47/+93
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When an MMP write is skipped, or fails, and time since mts->mmp_last_write is already greater than mts->mmp_delay, increase mts->mmp_delay. The original code only updated mts->mmp_delay when a write succeeded, but this results in the write(s) after delays and failed write(s) reporting an ub_mmp_delay which is too low. Update mmp_last_write and mmp_delay if a txg sync was successful. At least one uberblock was written, thus extending the time we can be sure the pool will not be imported by another host. Do not allow mmp_delay to go below (MSEC2NSEC(zfs_multihost_interval) / vdev_count_leaves()) so that a period of frequent successful MMP writes, e.g. due to frequent txg syncs, does not result in an import activity check so short it is not reliable based on mmp thread writes alone. Remove unnecessary local variable, start. We do not use the start time of the loop iteration. Add a debug message in spa_activity_check() to allow verification of the import_delay value and to prove the activity check occurred. Alter the tests that import pools and attempt to detect an activity check. Calculate the expected duration of spa_activity_check() based on module parameters at the time the import is performed, rather than a fixed time set in mmp.cfg. The fixed time may be wrong. Also, use the default zfs_multihost_interval value so the activity check is longer and easier to recognize. Reviewed-by: Tony Hutter <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: Olaf Faaland <[email protected]> Closes #7330
* Fedora 28: Fix misc bounds check compiler warningsTony Hutter2018-04-0410-34/+77
| | | | | | | | | Fix a bunch of (mostly) sprintf/snprintf truncation compiler warnings that show up on Fedora 28 (GCC 8.0.1). Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tony Hutter <[email protected]> Closes #7361 Closes #7368
* Fix spa reference leak in zfs_ioc_pool_scanLOLi2018-04-031-3/+3
| | | | | | | | | | | | zfs_ioc_pool_scan leaks a spa reference when zc->zc_flags is not a valid pool_scrub_cmd_t: this could happen if the userland binaries and ZFS kernel module differ in version and would prevent the pool from being exported. Reviewed by: Matt Ahrens <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: loli10K <[email protected]> Closes #7380
* Fix add_nested_replacing_spare test caseLOLi2018-04-031-1/+1
| | | | | | | | | | | | | | | | | | | | | | | Use 'zpool reopen' instead of 'zpool scrub' to kick in the spare device: this is required to avoid spurious failures caused by a race condition in events processing by the ZFS Event Daemon: P1 (zpool scrub) P2 (zed) --- zfs_ioc_pool_scan() -> dsl_scan() -> vdev_reopen() -> vdev_set_state(VDEV_STATE_CANT_OPEN) zfs_ioc_vdev_attach() -> spa_vdev_attach() -> dsl_resilver_restart() -> dsl_sync_task() -> dsl_scan_setup_check() <- dsl_scan_setup_check(): EBUSY Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: loli10K <[email protected]> Closes #7247 Closes #7342
* Remove ASSERT() in l2arc_apply_transforms()Tim Chase2018-03-311-4/+4
| | | | | | | | | | The ASSERT was erroneously copied from the next section of code. The buffer's size should be expanded from "psize" to "asize" if necessary. Reviewed-by: Tom Caputi <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tim Chase <[email protected]> Closes #7375
* Decryption error handling improvementsTom Caputi2018-03-319-23/+68
| | | | | | | | | | | | | | | | | | Currently, the decryption and block authentication code in the ZIO / ARC layers is a bit inconsistent with regards to the ereports that are produces and the error codes that are passed to calling functions. This patch ensures that all of these errors (which begin as ECKSUM) are converted to EIO before they leave the ZIO or ARC layer and that ereports are correctly generated on each decryption / authentication failure. In addition, this patch fixes a bug in zio_decrypt() where ECKSUM never gets written to zio->io_error. Reviewed by: Matt Ahrens <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7372
* Encrypted dnode blocks should be prefetched rawTom Caputi2018-03-311-2/+7
| | | | | | | | | | | | | | | Encrypted dnode blocks are always initially read as raw data and converted to decrypted data when an encrypted bonus buffer is needed. This allows the DMU to be used for things like fetching the DMU master node without requiring keys to be loaded. However, dbuf_issue_final_prefetch() does not currently read the data as raw. The end result of this is that prefetched dnode blocks are read twice from disk: once decrypted and then again as raw data. This patch corrects the issue by adding the flag when appropriate. Reviewed by: Matt Ahrens <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7362
* Fix hung z_zvol tasks during 'zfs receive'LOLi2018-03-303-1/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | During a receive operation zvol_create_minors_impl() can wait needlessly for the prefetch thread because both share the same tasks queue. This results in hung tasks: <3>INFO: task z_zvol:5541 blocked for more than 120 seconds. <3> Tainted: P O 3.16.0-4-amd64 <3>"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. The first z_zvol:5541 (zvol_task_cb) is waiting for the long running traverse_prefetch_thread:260 root@linux:~# cat /proc/spl/taskq taskq act nthr spwn maxt pri mina spl_system_taskq/0 1 2 0 64 100 1 active: [260]traverse_prefetch_thread [zfs](0xffff88003347ae40) wait: 5541 spl_delay_taskq/0 0 1 0 4 100 1 delay: spa_deadman [zfs](0xffff880039924000) z_zvol/1 1 1 0 1 120 1 active: [5541]zvol_task_cb [zfs](0xffff88001fde6400) pend: zvol_task_cb [zfs](0xffff88001fde6800) This change adds a dedicated, per-pool, prefetch taskq to prevent the traverse code from monopolizing the global (and limited) system_taskq by inappropriately scheduling long running tasks on it. Reviewed-by: Albert Lee <[email protected]> Signed-off-by: Brian Behlendorf <[email protected]> Signed-off-by: loli10K <[email protected]> Closes #6330 Closes #6890 Closes #7343
* zfs-functions: skip lines where mntpnt is 'none'Georgy Yakovlev2018-03-301-0/+2
| | | | | | | | | | | | | | This fixes zfs-mount initscript trying to mount swap volumes as filesystems or anything that has 'none' as a mountpoint in /etc/fstab. Additionally, fixes trying to mount swap volumes as a filesystem on RHEL. RHEL defines mountpoint for swap as `swap`. Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: George Melikov <[email protected]> Signed-off-by: Georgy Yakovlev <[email protected]> Closes #7346 Closes #7347
* OpenZFS 9164 - assert: newds == os->os_dsl_datasetAndriy Gapon2018-03-306-25/+32
| | | | | | | | | | | | | | | | | | | | Authored by: Andriy Gapon <[email protected]> Reviewed by: Matt Ahrens <[email protected]> Reviewed by: Don Brady <[email protected]> Reviewed-by: loli10K <[email protected]> Reviewed-by: Tony Hutter <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Approved by: Richard Lowe <[email protected]> Ported-by: Giuseppe Di Natale <[email protected]> Porting Notes: * Re-enabled and tweaked the zpool_upgrade_007_pos test case to successfully run in under 5 minutes. OpenZFS-issue: https://www.illumos.org/issues/9164 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/0e776dc06a Closes #6112 Closes #7336
* Add support for nvme based devidsDon Brady2018-03-291-2/+13
| | | | | | | | | | | Adds a devid for nvme devices. This is very similar to how the other 'bus' (scsi|sata|usb) devids are generated. The devid resides in a name/value pair in the leaf vdevs in a zpool config. Reviewed-by: Tony Hutter <[email protected]> Reviewed-by: Richard Elling <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Don Brady <[email protected]> Closes #7356
* Resolve QAT issues with incompressible dataTom Caputi2018-03-293-37/+106
| | | | | | | | | | | | | | | | | | | | | | Currently, when ZFS wants to accelerate compression with QAT, it passes a destination buffer of the same size as the source buffer. Unfortunately, if the data is incompressible, QAT can actually "compress" the data to be larger than the source buffer. When this happens, the QAT driver will return a FAILED error code and print warnings to dmesg. This patch fixes these issues by providing the QAT driver with an additional buffer to work with so that even completely incompressible source data will not cause an overflow. This patch also resolves an error handling issue where incompressible data attempts compression twice: once by QAT and once in software. To fix this issue, a new (and fake) error code CPA_STATUS_INOMPRESSIBLE has been added so that the calling code can correctly account for the difference between a hardware failure and data that simply cannot be compressed. Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Weigang Li <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7338
* Fix ASSERT in dsl_scan_fini() and cleanup commentsTom Caputi2018-03-281-13/+13
| | | | | | | | | | | | | | | This patch fixes an issue where dsl_scan_prefetch_cb() might add more prefetch I/Os to the prefetch queue after prefetching has been completed. This was happening because that code was checking scn->scn_suspending instead of scn->scn_prefetch_stop. This occasionally triggered an ASSERT during ztest runs in dsl_scan_fini() when the code attempted to destroy an AVL tree that still had entires in it. This patch also includes a number of spelling corrections and comment cleanups throughout dsl_scan.c Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7353
* chmod -x on etc/init.d/zfs-*.in automake filesTony Hutter2018-03-284-0/+0
| | | | | | | | | | | | Clear executable bit on zfs-import.in, zfs-mount.in, zfs-share.in, and zfs-zed.in. These are automake files and should not be marked executable. This fixes a RPM build error on Fedora 28. Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tony Hutter <[email protected]> Closes #7355 Closes #7327
* Fix mmap / libaio deadlockBrian Behlendorf2018-03-2814-5/+184
| | | | | | | | | | | | | | Calling uiomove() in mappedread() under the page lock can result in a deadlock if the user space page needs to be faulted in. Resolve the issue by dropping the page lock before the uiomove(). The inode range lock protects against concurrent updates via zfs_read() and zfs_write(). Reviewed-by: Albert Lee <[email protected]> Reviewed-by: Chunwei Chen <[email protected]> Signed-off-by: Brian Behlendorf <[email protected]> Closes #7335 Closes #7339
* Remove libattr requirementDeHackEd2018-03-274-15/+2
| | | | | | | | | | RHEL/CentOS 6 supports sys/xattr.h eliminating the need for libattr-devel as a dependency. Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: DHE <[email protected]> Closes #7344 Closes #7351
* OpenZFS 9321 - arc_loan_compressed_buf() can increment arc_loaned_bytes by ↵Allan Jude2018-03-261-2/+2
| | | | | | | | | | | | | | | | the wrong value Authored by: Allan Jude <[email protected]> Reviewed by: Matt Ahrens <[email protected]> Reviewed by: George Wilson <[email protected]> Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Approved by: Garrett D'Amore <[email protected]> Ported-by: Giuseppe Di Natale <[email protected]> OpenZFS-issue: https://www.illumos.org/issues/9321 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/92b05f3a18 Closes #7333
* Fedora 28: Fix "Macro %_dracutdir has empty body"Tony Hutter2018-03-251-4/+17
| | | | | | | | | | | If you run ./configure --with-config=srpm, it will not trigger the user m4 scripts to populate the dracut and udev directories. This causes a build error on Fedora 28. Make the dracut and udev lines conditional to get around this. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tony Hutter <[email protected]> Closes #7326 Closes #7328
* Don't count embedded bps in read statsTom Caputi2018-03-231-1/+3
| | | | | | | | | | | | | | Currently, ZFS tracks statistics about calls to arc_read() via the /proc/spl/kstat/zfs/<pool>/reads file for debugging. Unfortunately, this file currently counts embedded bps as disk reads since they are technically processed by the ZIO layer. This pollutes the log since the ARC will never cache embedded bps. This patch corrects this issue by preventing the logging of embedded bp reads. Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7334
* OpenZFS 9193 - bootcfg -C doesn't workPaul Dagnelie2018-03-221-2/+3
| | | | | | | | | | | | | | | | | | | | When given an empty string as a rootds value, bootcfg -C fails with the error message 'could not set nextboot: '' is an invalid name'. This should be allowed because it represents clearing the nextboot configuration. Authored by: Paul Dagnelie <[email protected]> Reviewed by: Chris Williamson <[email protected]> Reviewed by: Sebastien Roy <[email protected]> Reviewed by: Igor Kozhukhov <[email protected]> Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Approved by: Robert Mustacchi <[email protected]> Ported-by: Giuseppe Di Natale <[email protected]> OpenZFS-issue: https://www.illumos.org/issues/9193 OpenZFS-commit: https://github.com/openzfs/openzfs/commit/504645d227 Closes #7230
* Clarify zpool actions for an intent log devicePeter Ashford2018-03-221-2/+3
| | | | | | | | | | Updated the "Intent Log" section of the "zpool" manual page to properly reflect the actions that may be performed. Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Peter Ashford <[email protected]> Closes #6938 Closes #7318
* modprobe zfs during dracut mountkpande2018-03-221-0/+1
| | | | | | | | | | | Resolves importing root pool during boot in dracut. This case was inadvertently broken with the module autoloading change in #7287. Reviewed-by: Matthew Thode <[email protected]> Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: Kash Pande <[email protected]> Closes #7322
* enable zfs_dbgmsg() by default, without dprintf()Matthew Ahrens2018-03-215-28/+38
| | | | | | | | | | | | | | | | | | | | | | | zfs_dbgmsg() should record a message by default. As a general principal, these messages shouldn't be too verbose. Furthermore, the amount of memory used is limited to 4MB (by default). dprintf() should only record a message if this is a debug build, and ZFS_DEBUG_DPRINTF is set in zfs_flags. This flag is not set by default (even on debug builds). These messages are extremely verbose, and sometimes nontrivial to compute. SET_ERROR() should only record a message if ZFS_DEBUG_SET_ERROR is set in zfs_flags. This flag is not set by default (even on debug builds). This brings our behavior in line with illumos. Note that the message format is unchanged (including file, line, and function, even though these are not recorded on illumos). Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Prakash Surya <[email protected]> Signed-off-by: Matthew Ahrens <[email protected]> Closes #7314
* Fix spelling errors in commentsTom Caputi2018-03-214-47/+48
| | | | | | | | | | This patch simply corrects some spelling / grammar errors in the QAT and encryption code comments. No functional changes Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7319
* Add support for nvme disk detectiontimor2018-03-211-0/+14
| | | | | | | | | | | | | This treats /dev/nvme.. devices the same way as /dev/sd... devices. The motivation behind this is that whole disk detection did not work on nvme SSDs without that, because it DKC_UNKNOWN was returned for such devices. Perhaps there should be a separate DKC_ type for this, but I don't know enough about the code to know the implications of that. Reviewed-by: Don Brady <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: timor <[email protected]> Closes #7304
* Add comments for portable dnode / objset flagsTom Caputi2018-03-202-1/+13
| | | | | | | | | | | This patch adds some comments describing the purpose of "portable" dnode and objset flags so that it is clear when new flags should be added to the repective flag masks. This patch includes no functional changes. Reviewed-by: George Melikov <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7313
* Add JSON output support to channel programsAlek P2018-03-1913-11/+639
| | | | | | | | | | | | | | | | | | | The changes piggyback JSON output support on top of channel programs (#6558). This way the JSON output support is targeted to scripting use cases and is easily maintainable since it really only touches one function (zfs_do_channel_program()). This patch ports Joyent's JSON nvlist library from illumos to enable easy JSON printing of channel program output nvlist. To keep the delta small I also took advantage of the fact that printing in zfs_do_channel_program() was almost always done before exiting the program. Reviewed by: Matt Ahrens <[email protected]> Reviewed-by: Tony Hutter <[email protected]> Reviewed-by: Richard Elling <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Alek Pinchuk <[email protected]> Closes #7281
* Fix deadlock in IO pipelineBrian Behlendorf2018-03-161-1/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In vdev_queue_aggregate() the zio_execute() bypass should not be called under the vdev queue lock. This can result in a deadlock as shown in the stack traces below. Drop the vdev queue lock then walk the parents of the aggregate IO to determine the list of component IOs to be bypassed. This can be done safely without holding the io_lock since the new aggregate IO has not yet been returned and its parents cannot change. --- THREAD 1 --- arc_read() zio_nowait() zio_vdev_io_start() vdev_queue_io() <--- mutex_enter(vq->vq_lock) vdev_queue_io_to_issue() vdev_queue_aggregate() zio_execute() zio_vdev_io_assess() zio_wait_for_children() <- mutex_enter(zio->io_lock) --- THREAD 2 --- (inverse order) arc_read() zio_change_priority() <- mutex_enter(zio->zio_lock) vdev_queue_change_io_priority() <- mutex_enter(vq->vq_lock) Reviewed-by: Tom Caputi <[email protected]> Reviewed-by: Don Brady <[email protected]> Signed-off-by: Brian Behlendorf <[email protected]> Closes #7307
* Allow QAT to handle non page-aligned buffersTom Caputi2018-03-162-34/+60
| | | | | | | | | | | | | | This patch adds some handling to the QAT acceleration functions that allows them to handle buffers that are not aligned with the page cache. At the moment this never happens since callers only happen to work with page-aligned buffers, but this code should prevent headaches if this isn't always true in the future. This patch also adds some cleanups to align the QAT compression code with the encryption and checksumming code. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Weigang Li <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7305
* Report pool suspended due to MMPOlaf Faaland2018-03-1510-11/+38
| | | | | | | | | | | | | | | | | | | | | | | | | | When the pool is suspended, record whether it was due to an I/O error or due to MMP writes failing to succeed within the required time. Change spa_suspended from uint8_t to zio_suspend_reason_t to store the reason. When userspace queries pool status via spa_tryimport(), report the reason the pool was suspended in a new key, ZPOOL_CONFIG_SUSPENDED_REASON. In libzfs, when interpreting the returned config nvlist, report suspension due to MMP with a new pool status enum value, ZPOOL_STATUS_IO_FAILURE_MMP. In status_callback(), which generates and emits the message when 'zpool status' is executed, add a case to print an appropriate message for the new pool status enum value. Reviewed-by: George Melikov <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Tony Hutter <[email protected]> Signed-off-by: Olaf Faaland <[email protected]> Closes #7296
* SHA256 QAT accelerationTom Caputi2018-03-155-27/+235
| | | | | | | | | | | | This patch enables acceleration of SHA256 checksums using Intel Quick Assist Technology. This patch also fixes up and refactors some of the code from QAT encryption to make the behavior consistent. Reviewed-by: Brian Behlendorf <[email protected]> Signed-off-by: Chengfeix Zhu <[email protected]> Signed-off-by: Weigang Li <[email protected]> Signed-off-by: Tom Caputi <[email protected]> Closes #7295
* OpenZFS 9076 - Adjust perf test concurrency settingsStephen Blinick2018-03-155-10/+55
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | ZFS Performance test concurrency should be lowered for better latency Work by Stephen Blinick. Nightly performance runs typically consist of two levels of concurrency; and both are fairly high. Since the IO runs are to a ZFS filesystem, within a zpool, which is based on some variable number of vdev's, the amount of IO driven to each device is variable. Additionally, different device types (HDD vs SSD, etc) can generally handle a different amount of concurrent IO before saturating. Nevertheless, in practice, it appears that most tests are well past the concurrency saturation point and therefore both perform with the same throughput, the maximum of the device. Because the queuedepth to the device(s) is so high however, the latency is much higher than the best possible at that throughput, and increases linearly with the increase in concurrency. This means that changes in code that impact latency during normal operation (before saturation) may not be apparent when a large component of the measured latency is from the IO sitting in a queue to be serviced. Therefore, changing the concurrency settings is recommended Authored by: Stephen Blinick <[email protected]> Reviewed-by: George Melikov <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Reviewed-by: Brian Behlendorf <[email protected]> Reviewed by: Dan Kimmel <[email protected]> Reviewed by: John Wren Kennedy <[email protected]> Ported-by: John Wren Kennedy <[email protected]> OpenZFS-issue: https://www.illumos.org/issues/9076 OpenZFS-commit: https://github.com/openzfs/openzfs/pull/562 Upstream bug: DLPX-45477 Closes #7302
* receive_spill does not byte swap spill contentsPaul Zuchowski2018-03-151-0/+9
| | | | | | | | | In zfs receive, the function receive_spill should account for spill block endian conversion as a defensive measure. Reviewed-by: Brian Behlendorf <[email protected]> Reviewed-by: Tom Caputi <[email protected]> Signed-off-by: Paul Zuchowski <[email protected]> Closes #7300
* OpenZFS 9188 - increase size of dbuf cache to reduce indirect block ↵Brian Behlendorf2018-03-132-11/+66
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | decompression With compressed ARC (bug 6950) we use up to 25% of our CPU to decompress indirect blocks, under a workload of random cached reads. To reduce this decompression cost, we would like to increase the size of the dbuf cache so that more indirect blocks can be stored uncompressed. If we are caching entire large files of recordsize=8K, the indirect blocks use 1/64th as much memory as the data blocks (assuming they have the same compression ratio). We suggest making the dbuf cache be 1/32nd of all memory, so that in this scenario we should be able to keep all the indirect blocks decompressed in the dbuf cache. (We want it to be more than the 1/64th that the indirect blocks would use because we need to cache other stuff in the dbuf cache as well.) In real world workloads, this won't help as dramatically as the example above, but we think it's still worth it because the risk of decreasing performance is low. The potential negative performance impact is that we will be slightly reducing the size of the ARC (by ~3%). Porting Notes: * Added modules options to zfs-module-parameters.5 man page. * Preserved scaling based on target ARC size rather than max ARC size. Authored by: George Wilson <[email protected]> Reviewed by: Dan Kimmel <[email protected]> Reviewed by: Prashanth Sreenivasa <[email protected]> Reviewed by: Paul Dagnelie <[email protected]> Reviewed-by: Richard Elling <[email protected]> Reviewed-by: George Melikov <[email protected]> Reviewed-by: Giuseppe Di Natale <[email protected]> Ported-by: Brian Behlendorf <[email protected]> OpenZFS-issue: https://www.illumos.org/issues/9188 OpenZFS-commit: https://github.com/openzfs/openzfs/pull/564 Upstream bug: DLPX-46942 Closes #7273