JFS

JFS - Journaled File System for OS/2 v1.09.10 Introduction ============ This package is an Installable File System which support the JFS file system. This package contains the following files: cachejfs.exe - query and set parameters of cache cstats.exe - query usage of cache extendfs.exe - for internal use jfs.ifs - IFS driver jfschk32.exe - chkdsk launcher during boot (can be used instead of chkdsk) ujfs.dll - used by utilities: format, sys, chkdsk jfsboot.exe - updates os2boot on JFS boot disk. Copyright and License ===================== Copyright (c) 2011-2023 David Azarewicz <david@88watts.net> All rights reserved. (c) Copyright IBM Corp. 1996, 1999 All rights reserved. Important! By downloading or using this software, you acknowledge that you have read and consent to the license agreement contained in the separate License.txt file, that you understand it, and that you agree to be bound by its terms. JFS.IFS Options =============== IFS=JFS.IFS <parameters> /CACHE:n - Specifies the size of the JFS buffer cache in Kilobytes. By default, this size is set to 12.5 percent of real memory. /L:OFF - Specifies the lazy write parameters in seconds. OFF forces all to be synchronous. (JFS ignores any characters between the L (or l) and the colon, so valid flags are /L: /LAZY: /LW: /lazywrite: etc.) /L:Synctime,Max,Min - Synctime is the interval at which the sync thread runs, default is 64. Max is the longest time that a modified file is kept in cache, default is synctime*4. Min is the time indicating a "recent" change. Changes newer than this value are not written unless the last write was older than Max, default is Min(1, synctime/8). /RO:drive[drive...] - Specifies a list of JFS file systems, identified by drive letter, to be mounted read only. /AUTOCHECK:drive[drive...] - Specifies a list of JFS file systems, identified by drive letter, to be included in automatic recovery by CHKDSK at the time of JFS initialization. If an asterisk (*) is specified instead of a drive letter list, all JFS file systems will be included in automatic recovery during JFS initialization. Normally CHKDSK will not touch a file system if it is marked clean. If the file system is marked dirty and the drive letter is preceded by a plus sign (+), CHKDSK will perform a full integrity check on the drive even if journal log replay has restored the drive to an apparently consistent state (forceifdirty). If any drive letter is preceded by an exclaimation point (!), CHKDSK will perform a full integrity check on the drive even if the file system is marked clean (forceifclean). Examples: /AUTOCHECK:+F* autochecks all JFS disks, and forces a full chkdsk on drive F if dirty. /A:!E* autochecks all JFS disks and forces a full chkdsk on drive E even if clean. /A:!C+D* autochecks all JFS disks, forces a full chkdsk on drive C even if clean, and forces a full chkdsk on drive D if dirty. /A:CD autochecks disks C and D only. Other disks are not included in the autocheck. Getting Support and Reporting Problems ====================================== For more information and to report problems please visit: https://www.arcanoae.com and click on SUPPORT. Or go directly to the JFS support wiki at: https://www.arcanoae.com/wiki/jfs/ Change Log ========== 1.09.10 - 30-Jun-2023 - David Azarewicz --------------------------------------- Fixed a logic error that caused a memory buffer to be released twice in the event a media write error occurred while writing an EA or ACL. 1.09.09 - 29-Nov-2020 - David Azarewicz --------------------------------------- Fixed the cache read-ahead problem that can cause a file system hang in some rare cases. The Cache Read-Ahead Hang occurs when the requested read size for a DosRead is greater than the cache size, and the remaining file data to read is greater than the cache size, and there is already at least one block of file data in the cache. The read-ahead algorithm will attempt to read-ahead more blocks than will fit in the cache and will hang. This can happen in a number of ways, but the most common way is to open the file twice and attempt to read more than the cache size in a single operation. For the default cache size, this size is about a 64 MB file. If the JFS cache size was changed by the /CACHE: option then the size required to produce the hang will vary according to the cache size specified. The cache read-ahead hang will not occur if there are no blocks from the file already in the cache. For example if the file is only opened once and immediately read. The hang also will not happen if the read size is less than the cache size, regardless of cache content, and regardless of the file size. 1.09.08 - 10-Jul-2020 - David Azarewicz --------------------------------------- Fixed a problem in JFSBOOT that causes an IPE when trying to access an empty directory during basedev load time. No change to the IFS. 1.09.07 - 10-Mar-2016 - David Azarewicz --------------------------------------- There are only cosmetic changes in the version. No code changes at all were made which affect disk operations. Changed command parser to add an option to call existing chkdsk functions differently. 1.09.06 - 23-Oct-2013 - David Azarewicz --------------------------------------- Fixed a defect in the IFS which caused incomplete log entries to be written. If the system crashed before the affected sectors were written to the disk, a subsequent log redo would corrupt the disk and mark the disk as clean. 1.09.05 - 24-Jan-2012 - David Azarewicz --------------------------------------- Added caching to the JFS boot loader to enhance boot performance.